Expedient knowledge corpus
Collection
Source-verified QA corpora behind the CTRL models • 8 items • Updated
Datasets:
id stringlengths 8 10 | vendor stringclasses 3
values | product stringclasses 102
values | category stringclasses 13
values | record_type stringclasses 8
values | intent_family stringlengths 10 53 ⌀ | version_scope stringclasses 64
values | question stringlengths 13 158 | answer stringlengths 89 630 | source stringclasses 266
values | evidence listlengths 1 9 ⌀ | review_status stringclasses 2
values | training_eligible bool 2
classes | rights_status stringclasses 2
values | execution_level stringclasses 2
values | collection stringclasses 6
values | difficulty stringclasses 3
values | repair_note stringclasses 2
values | availability_scope stringclasses 1
value | support_scope stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
expkb-0001 | expedient | AI CTRL | service-overview | reference | knowledge-base-navigation | KB page as retrieved 2026-10-31 | Where should I look for Expedient product guidance and support? | The Expedient Knowledge Base offers helpful articles and guides for a range of products and services. To find relevant guidance, search from the bar above or use the navigation pane. For more information about Expedient, the page directs visitors to expedient.com. Existing clients can also open a support ticket through... | https://kb.expedient.com/docs/welcome | [
{
"url": "https://kb.expedient.com/docs/welcome.md",
"excerpt": "Here you will find helpful articles and guides for many of our products and services. You can search from the bar above or use the navigation pane to find what you're looking for."
},
{
"url": "https://kb.expedient.com/docs/welcome.md"... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0002 | expedient | AI CTRL | service-overview | reference | ai-ctrl-platform-overview | KB page as retrieved 2026-10-31 | What is AI CTRL, and how is customer data handled? | Expedient describes AI CTRL as a security-first, multi-model AI gateway that provides access to models from OpenAI, Anthropic, Google, and Perplexity through one interface. Each organization has its own dedicated private environment. The guide also states that customer data is isolated and is not used to train models. | https://kb.expedient.com/docs/v3-user-guide | [
{
"url": "https://kb.expedient.com/docs/v3-user-guide.md",
"excerpt": "Expedient AI CTRL is a multi-model AI gateway built with a security-first approach. Instead of being locked into a single provider, your organization gains access to models from four major providers (OpenAI, Anthropic, Google, and Perple... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0003 | expedient | AI CTRL | how-to | how-to | sso-first-login | KB page as retrieved 2026-10-31 | How do I sign in to AI CTRL for the first time? | Use the SSO button on the login page and authenticate through your organization’s identity provider. The user guide says there is no separate AI CTRL password to manage. If you can authenticate with that provider, you can use that same sign-in flow for AI CTRL. | https://kb.expedient.com/docs/v3-user-guide | [
{
"url": "https://kb.expedient.com/docs/v3-user-guide.md",
"excerpt": "Use the single sign-on (SSO) button on the login page to authenticate through your organization's identity provider. If you can authenticate there, you can get in: there is no separate AI CTRL password to manage."
}
] | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0004 | expedient | AI CTRL | how-to | how-to | mobile-pwa-installation | KB page as retrieved 2026-10-31 | Can I install AI CTRL on my phone without an app-store download? | Yes. The chat app guide says there is nothing to download: adding the shortcut points to the secure address you already use, and you continue signing in the same way. On a phone, open AI CTRL while logged in, use the browser’s add-to-home-screen or install action, and confirm the app icon appears. | https://kb.expedient.com/docs/chat-app-installation | [
{
"url": "https://kb.expedient.com/docs/chat-app-installation.md",
"excerpt": "**There's nothing to download.** This creates a shortcut to the same secure address you already use, and you'll still sign in the same way. No app store, no install package, and nothing for IT to approve."
},
{
"url": "ht... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0005 | expedient | AI CTRL | how-to | how-to | desktop-web-app-installation | KB page as retrieved 2026-10-31 | How do I add AI CTRL as a desktop app in Chrome? | While logged into AI CTRL Chat, select the install icon on the right side of Chrome’s address bar, then choose **Install** in the popup. The guide says this creates a desktop shortcut. Open that shortcut to check it launches in its own window without browser tabs or a toolbar. | https://kb.expedient.com/docs/chat-app-installation | [
{
"url": "https://kb.expedient.com/docs/chat-app-installation.md",
"excerpt": "While logged into AI CTRL Chat, select the install icon on the right side of the address bar — a computer with a downward arrow."
},
{
"url": "https://kb.expedient.com/docs/chat-app-installation.md",
"excerpt": "In th... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0006 | expedient | AI CTRL | how-to | how-to | model-selection | KB page as retrieved 2026-10-31 | How do I choose a model for a new chat? | Open the model name at the top of the chat area to see the available list. Models include descriptions to help match them to tasks, but administrators determine what appears for each user. You can pin frequently used models, set a personal default, and switch models mid-conversation while keeping the conversation conte... | https://kb.expedient.com/docs/choosing-a-model | [
{
"url": "https://kb.expedient.com/docs/choosing-a-model.md",
"excerpt": "- **Pin the models you use** to keep them in the sidebar, a click away from any screen\n- **Set your default model** in your settings: this is what opens when you start a new chat\n- **Switch mid-conversation** whenever you want; ever... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0007 | expedient | AI CTRL | faq | faq | smart-router-vs-direct-model | KB page as retrieved 2026-10-31 | When should I use Smart Model Router instead of selecting a model directly? | Use Smart Model Router when you are unsure which of your custom models fits the request; it compares your message with custom-model descriptions. It does not choose among built-in models. If you already know the model you want, select it directly for that model’s output without depending on description matching. | https://kb.expedient.com/docs/choosing-a-model | [
{
"url": "https://kb.expedient.com/docs/choosing-a-model.md",
"excerpt": "If you're not sure which of your custom models fits what you're doing, select [**Smart Model Router**](https://kb.expedient.com/docs/smart-model-router-1) from the model list. It reads your message and routes it to the custom model (o... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0008 | expedient | AI CTRL | faq | reference | router-per-message-behavior | KB page as retrieved 2026-10-31 | Can Smart Model Router change models partway through one conversation? | Yes. The router evaluates each submitted message rather than choosing once for the whole conversation. It compares the message with the Description field of each available custom model and selects the closest match. As the conversation changes topics, different messages may therefore be routed to different custom model... | https://kb.expedient.com/docs/smart-model-router-1 | [
{
"url": "https://kb.expedient.com/docs/smart-model-router-1.md",
"excerpt": "The router re-evaluates on every message, not once per conversation."
},
{
"url": "https://kb.expedient.com/docs/smart-model-router-1.md",
"excerpt": "It compares your message against the Description field of each cust... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0009 | expedient | AI CTRL | faq | faq | side-by-side-model-comparison | KB page as retrieved 2026-10-31 | Does comparing multiple models use more of our resources? | Yes. AI CTRL sends the same prompt to several models and presents the responses in side-by-side swim lanes, but each lane is a separate request behind the scenes. The user guide recommends comparing two or three models at most; resource use increases proportionally with each model in the comparison. | https://kb.expedient.com/docs/v3-user-guide | [
{
"url": "https://kb.expedient.com/docs/v3-user-guide.md",
"excerpt": "You can send the same prompt to several models at once and view the responses in side-by-side \"swim lanes.\" A practical limit is two to three models; beyond that, the screen becomes cluttered and hard to read."
},
{
"url": "htt... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0010 | expedient | AI CTRL | faq | reference | model-type-task-fit | KB page as retrieved 2026-10-31 | Which model type should I use for routine questions versus complex analysis? | For general questions, drafting, summaries, rewrites, and quick lookups, the guide recommends trying an everyday or lightweight model first. Reasoning models are intended for multi-step problems, analysis, technical writing, and code. Large-context models fit long documents or large datasets, while specialized models h... | https://kb.expedient.com/docs/choosing-a-model | [
{
"url": "https://kb.expedient.com/docs/choosing-a-model.md",
"excerpt": "| Everyday / lightweight | General questions, drafting, summarizing, rewriting, quick lookups | Finishes first in a side-by-side, and sits at the low end of the ratecard. Handles more than people expect. Try it first |"
},
{
"... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0011 | expedient | AI CTRL | how-to | how-to | personal-system-prompt-setup | KB page as retrieved 2026-10-31 | How do I set a personal system prompt that applies to my chats? | Open your profile picture, choose **Settings**, select **General**, enter instructions in **System Prompt**, and save. Expedient says this is a personal setting that does not affect other users. The guide recommends describing who you are, what work you do, and how you want answers formatted and phrased. | https://kb.expedient.com/docs/personal-system-prompt | [
{
"url": "https://kb.expedient.com/docs/personal-system-prompt.md",
"excerpt": "1. Click your profile picture in the corner of the screen\n2. Open **Settings**\n3. Select **General**\n4. Enter your instructions in the **System Prompt** field\n5. Save"
},
{
"url": "https://kb.expedient.com/docs/perso... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0012 | expedient | AI CTRL | how-to | how-to | prompt-quality-practices | KB page as retrieved 2026-10-31 | How can I make my prompts more useful? | The AI basics article recommends specifying the desired outcome, giving relevant context or examples, breaking complex work into steps, and adjusting and retrying. It compares AI use to talking to a person: precise instructions and context make answers more helpful. For current or internal facts, provide sources or ena... | https://kb.expedient.com/docs/understanding-ai-basics | [
{
"url": "https://kb.expedient.com/docs/understanding-ai-basics.md",
"excerpt": "Using AI is a lot like talking to a person: the more precise your instructions and context, the more helpful the answer."
},
{
"url": "https://kb.expedient.com/docs/understanding-ai-basics.md",
"excerpt": "- Be spec... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0013 | expedient | AI CTRL | faq | faq | model-knowledge-cutoff | KB page as retrieved 2026-10-31 | Can AI CTRL answer questions about recent events or our internal processes without added sources? | Not reliably. Models have a knowledge cutoff and do not know events after their training ended; the article warns that answers about recent events can be confident but outdated. Models also do not know an organization’s internal projects, processes, or documents unless users provide them. The guide points to web search... | https://kb.expedient.com/docs/understanding-ai-basics | [
{
"url": "https://kb.expedient.com/docs/understanding-ai-basics.md",
"excerpt": "Every model has a knowledge cutoff date and knows nothing that happened after its training ended."
},
{
"url": "https://kb.expedient.com/docs/understanding-ai-basics.md",
"excerpt": "A model may have general knowled... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0014 | expedient | AI CTRL | how-to | how-to | attach-files-and-webpages | KB page as retrieved 2026-10-31 | How can I add a document or webpage to a conversation? | Use the **+** button at the bottom left of the chat window. You can select one or several files from a computer or phone, use the device camera to capture text, or paste a webpage URL so its text is brought into the conversation. Clear, straight-on images work best; handwriting, blurry photos, and complex layouts can b... | https://kb.expedient.com/docs/attaching-documents-data-to-conversations | [
{
"url": "https://kb.expedient.com/docs/attaching-documents-data-to-conversations.md",
"excerpt": "Most of what you'll attach comes from the **+ button** at the bottom left of the chat window."
},
{
"url": "https://kb.expedient.com/docs/attaching-documents-data-to-conversations.md",
"excerpt": "... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0015 | expedient | AI CTRL | faq | faq | webpage-attachment-scope | KB page as retrieved 2026-10-31 | If I attach one webpage, will AI CTRL crawl the rest of that site? | No. The webpage attachment reads the single page whose URL you provide; it does not crawl links or discover subpages, and wildcard URLs return an error. If you need a site’s content to be available regularly, the article says to discuss a data connector instead of relying on a one-page attachment. | https://kb.expedient.com/docs/attaching-documents-data-to-conversations | [
{
"url": "https://kb.expedient.com/docs/attaching-documents-data-to-conversations.md",
"excerpt": "This reads the single page you give it. It doesn't crawl links or discover subpages, and a wildcard URL will just error."
},
{
"url": "https://kb.expedient.com/docs/attaching-documents-data-to-conversa... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0016 | expedient | AI CTRL | how-to | how-to | attach-knowledge-notes-and-chats | KB page as retrieved 2026-10-31 | How do I reuse notes, shared knowledge, or an earlier chat as context? | AI CTRL lets you attach your own notes, curated organization-wide knowledge collections, or reference chats with their attached content. Type **#** in the message box to search the collections available to you; only collections you have access to appear. If a colleague’s collection is missing, ask its maintainer about ... | https://kb.expedient.com/docs/attaching-documents-data-to-conversations | [
{
"url": "https://kb.expedient.com/docs/attaching-documents-data-to-conversations.md",
"excerpt": "- **Attach Notes**: your own scratchpad. Good for text you reference repeatedly: a standing brief, a list, context you'd otherwise retype."
},
{
"url": "https://kb.expedient.com/docs/attaching-document... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0017 | expedient | AI CTRL | how-to | how-to | spreadsheet-analysis-tool | KB page as retrieved 2026-10-31 | How should I analyze an Excel or CSV file without losing its table structure? | Turn on the Structured Data Tool in the tools menu before asking spreadsheet questions. Expedient explains that ordinary text conversion can separate column headers from numbers and lead the model to guess. The tool loads the file into a table it can query, so calculations and comparisons are computed rather than estim... | https://kb.expedient.com/docs/attaching-documents-data-to-conversations | [
{
"url": "https://kb.expedient.com/docs/attaching-documents-data-to-conversations.md",
"excerpt": "For real spreadsheet work, turn on the **Structured Data Tool** in the tools menu."
},
{
"url": "https://kb.expedient.com/docs/attaching-documents-data-to-conversations.md",
"excerpt": "It loads th... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0018 | expedient | AI CTRL | faq | faq | verify-grounding-citations | KB page as retrieved 2026-10-31 | How do I check whether an answer came from an attached document? | When an answer uses an attached document or knowledge collection, AI CTRL includes citations. Click a citation to see the exact passage it came from. Expedient recommends checking citations before acting, because they help confirm the answer is grounded in your supplied material rather than assembled from general model... | https://kb.expedient.com/docs/working-with-responses | [
{
"url": "https://kb.expedient.com/docs/working-with-responses.md",
"excerpt": "When an answer draws on documents you attached or a knowledge collection, it includes citations. Click one to see the exact passage it came from."
},
{
"url": "https://kb.expedient.com/docs/working-with-responses.md",
... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0019 | expedient | AI CTRL | how-to | how-to | refine-and-manage-response-versions | KB page as retrieved 2026-10-31 | How can I revise an answer without losing the earlier version? | Use Regenerate to try again, add detail, make the response more concise, or request a change; each action creates a new version instead of overwriting the last. You can page between versions and delete weaker ones. The article notes that versions left in the conversation remain part of its context, so removing unwanted... | https://kb.expedient.com/docs/working-with-responses | [
{
"url": "https://kb.expedient.com/docs/working-with-responses.md",
"excerpt": "Each one produces a new version rather than overwriting the last."
},
{
"url": "https://kb.expedient.com/docs/working-with-responses.md",
"excerpt": "A version you leave behind stays in the conversation's context, wh... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0020 | expedient | AI CTRL | how-to | how-to | export-response-formats | KB page as retrieved 2026-10-31 | What do the PDF, Word, and Excel response export buttons include? | Use the export buttons on the response rather than asking the model for a download link. PDF and Word contain the response but not your prompt, so copy the question separately if you need it included. Excel exports table data only; narrative text around a table is not included. | https://kb.expedient.com/docs/working-with-responses | [
{
"url": "https://kb.expedient.com/docs/working-with-responses.md",
"excerpt": "Use the **export buttons** on the response: PDF, Word, or Excel for tables."
},
{
"url": "https://kb.expedient.com/docs/working-with-responses.md",
"excerpt": "**PDF and Word contain the response, not your prompt.** ... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0021 | expedient | AI CTRL | how-to | how-to | folder-context-inheritance | KB page as retrieved 2026-10-31 | Can I make a folder automatically provide instructions and reference files to its chats? | Yes. A folder can hold a system prompt and reference files, and conversations started inside it inherit that context. Subfolders inherit the parent’s context and can add their own instructions or materials. Expedient describes folders as personal, not shareable; use a knowledge collection when multiple people need the ... | https://kb.expedient.com/docs/folders-notes-sharing | [
{
"url": "https://kb.expedient.com/docs/folders-notes-sharing.md",
"excerpt": "A folder holds conversations, but it can also hold a **system prompt and reference files**, and every conversation you start inside it inherits them automatically."
},
{
"url": "https://kb.expedient.com/docs/folders-notes... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0022 | expedient | AI CTRL | faq | reference | notes-access-controls | KB page as retrieved 2026-10-31 | Who can see a note I create in AI CTRL? | Notes have Private, Shared, and Public access levels. Private notes are visible to you and platform administrators; Shared notes are visible to the people or groups you name; Public notes are visible to everyone in your organization. The note’s editability is configured separately, so access level and permission to edi... | https://kb.expedient.com/docs/folders-notes-sharing | [
{
"url": "https://kb.expedient.com/docs/folders-notes-sharing.md",
"excerpt": "| Private | You, plus platform administrators |\n| Shared | The people or groups you name |\n| Public | Everyone in your organization |"
},
{
"url": "https://kb.expedient.com/docs/folders-notes-sharing.md",
"excerpt":... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0023 | expedient | AI CTRL | faq | faq | shared-chat-snapshot | KB page as retrieved 2026-10-31 | What does someone receive when I share a conversation? | The recipient receives a read-only, point-in-time snapshot. They can read it or clone it into their own interactive copy, but their work does not change your original. The link does not update when you continue the conversation; share again to provide a newer snapshot. Unsharing revokes the existing link immediately. | https://kb.expedient.com/docs/folders-notes-sharing | [
{
"url": "https://kb.expedient.com/docs/folders-notes-sharing.md",
"excerpt": "What the recipient gets is a read-only snapshot."
},
{
"url": "https://kb.expedient.com/docs/folders-notes-sharing.md",
"excerpt": "They can read it, and they can clone it into their own copy and keep working from the... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0024 | expedient | AI CTRL | policy | policy | front-end-deletion-vs-audit-record | KB page as retrieved 2026-10-31 | If I delete a conversation, is its compliance record deleted too? | No. Deleting a conversation clears it from the front end and from front-end analytics, but the organization’s backend keeps a permanent log of prompts and responses. Expedient states that a user’s actions do not affect that compliance record. Treat front-end deletion as removing the conversation from your view, not as ... | https://kb.expedient.com/docs/settings-memory-your-data | [
{
"url": "https://kb.expedient.com/docs/settings-memory-your-data.md",
"excerpt": "Deleting a conversation clears it from the front end: your view, and the front-end analytics. **It does not remove it from the compliance record.**"
},
{
"url": "https://kb.expedient.com/docs/settings-memory-your-data... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0025 | expedient | AI CTRL | faq | faq | manual-memory-control | KB page as retrieved 2026-10-31 | Does AI CTRL automatically create memories about me? | The settings guide says memory is manual: nothing is created or deleted on its own. You can add a memory in Settings or ask the system to capture something during a conversation. Expedient distinguishes durable facts that apply across conversations from ordinary context, which stays within the current conversation. Adm... | https://kb.expedient.com/docs/settings-memory-your-data | [
{
"url": "https://kb.expedient.com/docs/settings-memory-your-data.md",
"excerpt": "Nothing is created or deleted on its own."
},
{
"url": "https://kb.expedient.com/docs/settings-memory-your-data.md",
"excerpt": "You can add a memory directly in Settings, or say *\"create a memory of this\"* mid-... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0026 | expedient | AI CTRL | how-to | how-to | personal-data-export-and-management | KB page as retrieved 2026-10-31 | Where can I export or review my AI CTRL data? | Open your profile picture and choose **Settings** to reach account settings. The data controls let you export everything as a single file, archive or delete everything, and review shared or archived conversations. The File Manager lets you see and remove uploaded files. Deleting conversations from the front end does no... | https://kb.expedient.com/docs/settings-memory-your-data | [
{
"url": "https://kb.expedient.com/docs/settings-memory-your-data.md",
"excerpt": "Everything personal to your account lives in one place: your preferences, what the AI remembers about you, and full control over your own content."
},
{
"url": "https://kb.expedient.com/docs/settings-memory-your-data.... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0027 | expedient | AI CTRL | how-to | how-to | third-party-chat-import-format | KB page as retrieved 2026-10-31 | What file should I prepare to import ChatGPT or Claude conversation history? | For ChatGPT, extract the export archive and import `conversations.json` directly; AI CTRL detects the standard export and converts it automatically. For Claude, extract the export archive, use the manifest to obtain `conversations-000.zip`, then extract that file to get `conversations.json`. AI CTRL does not convert Cl... | https://kb.expedient.com/docs/importing-chats-from-chatgpt-or-claude | [
{
"url": "https://kb.expedient.com/docs/importing-chats-from-chatgpt-or-claude.md",
"excerpt": "ChatGPT/OpenAI | No | Extract the export archive and import the `conversations.json` file directly."
},
{
"url": "https://kb.expedient.com/docs/importing-chats-from-chatgpt-or-claude.md",
"excerpt": "... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0028 | expedient | AI CTRL | faq | faq | chat-import-content-limitations | KB page as retrieved 2026-10-31 | Will importing a chat history preserve attachments and all platform-specific features? | No. Expedient cautions that bulk history import does not reliably bring in attachments, images, Projects, Memory, project instructions, Custom GPT settings, artifacts, tool outputs, platform-specific metadata, or some formatting and branches. If you need particular files available for reference, the article recommends ... | https://kb.expedient.com/docs/importing-chats-from-chatgpt-or-claude | [
{
"url": "https://kb.expedient.com/docs/importing-chats-from-chatgpt-or-claude.md",
"excerpt": "This process imports conversation history. It does not reliably import attachments, images, Claude Projects, Claude Memory, project instructions, Custom GPT settings, artifacts and tool outputs, platform-specific... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0029 | expedient | AI CTRL | release-notes | reference | release-cadence-and-highlights | KB page as retrieved 2026-10-31 | How often does AI CTRL change, and what are examples of recent platform updates? | The user guide describes a roughly two-week release cadence. The 0.10.2 notes highlight automatic context compaction, a reworked memory system, streamed reasoning, faster search and rendering, more granular administrator permissions, and richer usage analytics. The notes also state that self-service SSO is disabled acr... | https://kb.expedient.com/docs/v3-user-guide | [
{
"url": "https://kb.expedient.com/docs/v3-user-guide.md",
"excerpt": "AI CTRL evolves on a roughly two-week release cadence, so the available models and some minor features may change over time."
},
{
"url": "https://kb.expedient.com/docs/ai-ctrl-0-10-2-release-notes.md",
"excerpt": "- **Automa... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0030 | expedient | AI CTRL | faq | reference | administrator-roles-and-promotion | KB page as retrieved 2026-10-31 | What are the AI CTRL user roles, and how does someone become an administrator? | AI CTRL has User, Admin, and Pending roles. User is standard everyday access, Admin has full system access, and Pending freezes functionality. An existing administrator grants admin access by opening the user’s profile, changing the role to Admin, and saving; the user then refreshes to see the admin panel. | https://kb.expedient.com/docs/admin-guide | [
{
"url": "https://kb.expedient.com/docs/admin-guide.md",
"excerpt": "AI CTRL has three roles:"
},
{
"url": "https://kb.expedient.com/docs/admin-guide.md",
"excerpt": "- **User:** standard, everyday access. You decide which features and models each user receives.\n- **Admin:** full system access,... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0031 | expedient | AI CTRL | faq | reference | group-permission-inheritance | KB page as retrieved 2026-10-31 | If a user belongs to several groups, how are conflicting permissions resolved? | Group-based permissions layer on top of default permissions. For users who belong to multiple groups, AI CTRL applies the most permissive setting available across those groups. As a result, if any one group grants access to a model or feature, the user receives that access. Review group membership and permission grants... | https://kb.expedient.com/docs/admin-guide | [
{
"url": "https://kb.expedient.com/docs/admin-guide.md",
"excerpt": "**Group-based permissions** layer on top for elevated or specialized access."
},
{
"url": "https://kb.expedient.com/docs/admin-guide.md",
"excerpt": "**Permissions are additive at the highest permissible level.** A user in mult... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0032 | expedient | AI CTRL | policy | policy | least-privilege-defaults | KB page as retrieved 2026-10-31 | What default permissions does Expedient recommend for a new AI CTRL rollout? | The administrator guide recommends leaving Workspace creation off by default unless broad self-service model building fits the rollout. It also recommends keeping API key generation off by default and enabling it only for groups that need programmatic access. For chat deletion, consider restricting it to prevent accide... | https://kb.expedient.com/docs/admin-guide | [
{
"url": "https://kb.expedient.com/docs/admin-guide.md",
"excerpt": "- **Leave Workspace creation off by default** unless broad self-service model-building fits your rollout."
},
{
"url": "https://kb.expedient.com/docs/admin-guide.md",
"excerpt": "- **Leave API key generation off by default;** e... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0033 | expedient | AI CTRL | how-to | how-to | hide-retired-models-safely | KB page as retrieved 2026-10-31 | How should I remove an outdated model from the selector without breaking custom models? | Use **Hide** rather than making an outdated model private. Hiding removes it from the selector while keeping it reachable, so custom models built on that base can continue working. To hide one, open **Admin Panel → Settings → Models**, use the row’s ellipsis, and select **Hide**. Expedient warns that restricting the ba... | https://kb.expedient.com/docs/model-best-practices | [
{
"url": "https://kb.expedient.com/docs/model-best-practices.md",
"excerpt": "Use Hide rather than making an outdated model private. Hiding removes a model from the selector for everyone but keeps it reachable, so any custom model built on it keeps working. Restricting the model instead breaks those custom ... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0034 | expedient | AI CTRL | faq | faq | custom-model-base-access | KB page as retrieved 2026-10-31 | Why can a colleague not use a custom model that I shared with them? | A custom model inherits the access requirements of its base model. If the custom model is shared publicly but its base model is private, users without permission for that base will not be able to use the custom model. Align base-model and custom-model access, then grant the appropriate group or users access to the base... | https://kb.expedient.com/docs/admin-guide | [
{
"url": "https://kb.expedient.com/docs/admin-guide.md",
"excerpt": "A custom model inherits the access requirements of its base model."
},
{
"url": "https://kb.expedient.com/docs/admin-guide.md",
"excerpt": "If you build a custom model on a private base model but publish the custom model to eve... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0035 | expedient | AI CTRL | how-to | how-to | entra-sso-app-registration | KB page as retrieved 2026-10-31 | What are the key Entra app-registration settings for AI CTRL SSO? | The Entra guide requires at least Application Developer privileges. Create a web app registration, choose whether the tenant is single- or multi-tenant, and add the AI CTRL redirect URI using your confirmed gateway address. The guide also says to enable ID tokens for the hybrid flow and configure the front-channel logo... | https://kb.expedient.com/docs/microsoft-entra-id-configuration-for-sso | [
{
"url": "https://kb.expedient.com/docs/microsoft-entra-id-configuration-for-sso.md",
"excerpt": "You'll need at least **Application Developer** privileges in Entra."
},
{
"url": "https://kb.expedient.com/docs/microsoft-entra-id-configuration-for-sso.md",
"excerpt": "**Supported account types** ... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0036 | expedient | AI CTRL | troubleshooting | how-to | entra-readable-group-names | KB page as retrieved 2026-10-31 | Entra groups appear as GUIDs in AI CTRL. What setting controls readable group names? | The group-name sync guide says `cloud_displayname` works only when `groupMembershipClaims` is exactly `ApplicationGroup`; values such as `SecurityGroup` or `All` cause names to revert to GUIDs. The setup also requires adding the `cloud_displayname` property to both token types in the app manifest. Expedient must apply ... | https://kb.expedient.com/docs/microsoft-entra-id-group-name-sync | [
{
"url": "https://kb.expedient.com/docs/microsoft-entra-id-group-name-sync.md",
"excerpt": "The `cloud_displayname` property only functions when `groupMembershipClaims` is set to exactly **`ApplicationGroup`**. Any other value — including `SecurityGroup` or `All` — breaks the configuration and reverts group... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0037 | expedient | AI CTRL | faq | faq | entra-group-sync-source-of-truth | KB page as retrieved 2026-10-31 | Will manually assigned AI CTRL groups remain if Entra group sync is enabled? | For standard users, Entra is the source of truth at each login when OAuth group management is enabled. Memberships matching OAuth claims are added, and memberships not present in those claims—including manual AI CTRL assignments—are removed. AI CTRL administrator accounts bypass this sync; their group assignments must ... | https://kb.expedient.com/docs/microsoft-entra-id-group-name-sync | [
{
"url": "https://kb.expedient.com/docs/microsoft-entra-id-group-name-sync.md",
"excerpt": "When OAuth group management is enabled, a user's AI CTRL group memberships are synchronized with their OAuth claims each time they sign in."
},
{
"url": "https://kb.expedient.com/docs/microsoft-entra-id-group... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0038 | expedient | AI CTRL | how-to | how-to | okta-oidc-sso-setup | KB page as retrieved 2026-10-31 | How do I configure the Okta OIDC application for AI CTRL sign-in? | In Okta Admin Console, create an app integration with sign-in method **OIDC – OpenID Connect** and application type **Web Application**. Use the Authorization Code grant, enter the AI CTRL gateway’s `/oauth/oidc/callback` sign-in redirect and `/logout` sign-out redirect, then assign the users or groups who need access.... | https://kb.expedient.com/docs/okta-oidc-configuration-for-sso | [
{
"url": "https://kb.expedient.com/docs/okta-oidc-configuration-for-sso.md",
"excerpt": "- **Sign-in method** — select **OIDC – OpenID Connect**"
},
{
"url": "https://kb.expedient.com/docs/okta-oidc-configuration-for-sso.md",
"excerpt": "- **Application type** — select **Web Application**, then ... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0039 | expedient | AI CTRL | how-to | how-to | duo-oidc-sso-setup | KB page as retrieved 2026-10-31 | What does Duo require for an AI CTRL OIDC application? | In the Duo Admin Panel, add a **Generic OIDC Relying Party** application or the specific service you are integrating. Configure a clear application name, usually use the Authorization Code grant, and enter the AI CTRL sign-in callback URI. Select the required `openid` scope, optionally add `email` and `profile`, and en... | https://kb.expedient.com/docs/duo-oidc-configuration-for-sso | [
{
"url": "https://kb.expedient.com/docs/duo-oidc-configuration-for-sso.md",
"excerpt": "Search for **Generic OIDC Relying Party**, or the specific service you want to integrate."
},
{
"url": "https://kb.expedient.com/docs/duo-oidc-configuration-for-sso.md",
"excerpt": "- **Grant type** — usually... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0040 | expedient | AI CTRL | how-to | how-to | enable-api-key-generation | KB page as retrieved 2026-10-31 | Why can a group not create API keys even after I granted permission? | The global API key setting is above group permissions. If it is off, group permission does not enable key creation. First go to **Admin Panel → Settings → Authentication**, turn on **Enable API Keys**, and save. Then grant the API key permission to a named group and add only the users who need programmatic access. | https://kb.expedient.com/docs/configuring-api-access | [
{
"url": "https://kb.expedient.com/docs/configuring-api-access.md",
"excerpt": "The global setting sits above group permissions."
},
{
"url": "https://kb.expedient.com/docs/configuring-api-access.md",
"excerpt": "If it's off, granting a group the API key permission does nothing, with no error ex... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0041 | expedient | AI CTRL | policy | policy | api-consumption-and-fair-use | KB page as retrieved 2026-10-31 | Is API traffic included in the same fair-use allowance as interactive chat? | No. The API access guide says front-end chat falls under fair use, while API activity is metered against the organization’s consumption. It also notes that scripts do not have the natural pacing of a person typing. Treat enabling API access as both a security decision and a consumption decision, and coordinate with the... | https://kb.expedient.com/docs/configuring-api-access | [
{
"url": "https://kb.expedient.com/docs/configuring-api-access.md",
"excerpt": "**API usage is billed separately.** Front-end chat falls under fair use. Anything running through the API is metered against your organization's consumption, and a script has no natural pace limit the way a person typing does."
... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0042 | expedient | AI CTRL | policy | policy | api-key-account-scope-and-storage | KB page as retrieved 2026-10-31 | What should I know before storing or sharing an AI CTRL API key? | Each user creates their own key after access is granted. The key authenticates as that user and carries that user’s model access, so sharing it also shares that access. The guide warns that many tools store keys as plain text in configuration files, environment variables, or scripts; do not commit a key to a repository... | https://kb.expedient.com/docs/admin-guide | [
{
"url": "https://kb.expedient.com/docs/admin-guide.md",
"excerpt": "A custom model inherits the access requirements of its base model."
},
{
"url": "https://kb.expedient.com/docs/admin-guide.md",
"excerpt": "If you build a custom model on a private base model but publish the custom model to eve... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0043 | expedient | AI CTRL | faq | faq | api-interface-feature-scope | KB page as retrieved 2026-10-31 | Does API access include AI CTRL knowledge collections and prompt templates? | No. The API guide says an API-connected tool gets models and completions. Knowledge collections, the structured data tool, and prompt templates are features of the AI CTRL interface and remain there. Model access also follows the user associated with the key, so restricted models remain unavailable to users who lack th... | https://kb.expedient.com/docs/configuring-api-access | [
{
"url": "https://kb.expedient.com/docs/configuring-api-access.md",
"excerpt": "A tool using the API gets models and completions."
},
{
"url": "https://kb.expedient.com/docs/configuring-api-access.md",
"excerpt": "Knowledge collections, the structured data tool, and prompt templates are built in... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0044 | expedient | AI CTRL | how-to | how-to | mcp-server-registration | KB page as retrieved 2026-10-31 | How does an administrator add an MCP server connection to AI CTRL? | An AI CTRL Administrator using Open WebUI v0.9.6 or later can open **Admin Panel → Settings → Integrations** and add a connection. The article specifies **MCP Streamable HTTP** and **OAuth 2.1** among the configuration fields. Selecting **Register Client** triggers dynamic client registration; the client information is... | https://kb.expedient.com/docs/ai-ctrl-mcp-integration-configuration | [
{
"url": "https://kb.expedient.com/docs/ai-ctrl-mcp-integration-configuration.md",
"excerpt": "- **Role:** AI CTRL Administrator"
},
{
"url": "https://kb.expedient.com/docs/ai-ctrl-mcp-integration-configuration.md",
"excerpt": "- **Open WebUI Version:** v0.9.6 or later"
},
{
"url": "http... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0045 | expedient | AI CTRL | how-to | how-to | mcp-per-user-authorization | KB page as retrieved 2026-10-31 | After an MCP server is configured, can everyone use it immediately? | No. The MCP integration guide uses per-user OAuth, so each user authorizes their own account. In a chat, the user opens **Integrations → Tools**, enables the relevant MCP tool, completes the destination’s OAuth login, and accepts the authorization request. The user is redirected back to AI CTRL, and the tool is active ... | https://kb.expedient.com/docs/ai-ctrl-mcp-integration-configuration | [
{
"url": "https://kb.expedient.com/docs/ai-ctrl-mcp-integration-configuration.md",
"excerpt": "The tool uses **per-user OAuth** — each user must authorize their own account before using the tool."
},
{
"url": "https://kb.expedient.com/docs/ai-ctrl-mcp-integration-configuration.md",
"excerpt": "C... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0046 | expedient | AI CTRL | service-overview | reference | pii-filter-detection-types | KB page as retrieved 2026-10-31 | What types of PII can AI CTRL’s prompt filter detect and redact? | The PII filter article lists email addresses, phone numbers, Social Security numbers, credit card numbers, and IP addresses. It inspects text entered directly into a Secure Chat prompt and replaces a detected value with a redacted placeholder. The alerting platform can generate alerts when a redacted entry appears. | https://kb.expedient.com/docs/pii-filter-and-alerting | [
{
"url": "https://kb.expedient.com/docs/pii-filter-and-alerting.md",
"excerpt": "The filter supports the following PII types:"
},
{
"url": "https://kb.expedient.com/docs/pii-filter-and-alerting.md",
"excerpt": "- Email addresses\n- Phone numbers\n- Social Security Numbers (SSNs)\n- Credit card n... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0047 | expedient | AI CTRL | policy | policy | pii-filter-document-limitation | KB page as retrieved 2026-10-31 | Does the PII filter scan attached files or connected-source documents? | No. The filter does not scan documents added to a conversation, whether they are direct chat attachments or content retrieved from a connected source such as SharePoint. PII in those documents is not redacted and does not trigger a PII alert. Expedient advises removing or redacting PII that must not be exposed before m... | https://kb.expedient.com/docs/pii-filter-and-alerting | [
{
"url": "https://kb.expedient.com/docs/pii-filter-and-alerting.md",
"excerpt": "The PII filter does not scan the contents of documents added to a conversation, including files attached directly to a chat or content retrieved from a connected data source such as SharePoint."
},
{
"url": "https://kb.... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0048 | expedient | AI CTRL | how-to | how-to | configure-pii-filter-scope | KB page as retrieved 2026-10-31 | How can I apply the PII filter globally or only to selected models? | An administrator can enable the filter in **Admin Panel → Functions** and use the three-dot menu beside the PII Filter toggle to choose its scope. Leave **Global** enabled to apply it to all models; turn Global off to configure selected base models or custom models. Expedient advises copying the deployed function for a... | https://kb.expedient.com/docs/pii-filter-and-alerting | [
{
"url": "https://kb.expedient.com/docs/pii-filter-and-alerting.md",
"excerpt": "Locate **PII Filter** and confirm it is toggled on."
},
{
"url": "https://kb.expedient.com/docs/pii-filter-and-alerting.md",
"excerpt": "Choose how the filter is applied: leave **Global** enabled to apply it to all ... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0049 | expedient | AI CTRL | faq | reference | prompt-alert-event-behavior | KB page as retrieved 2026-10-31 | Does prompt monitoring create one alert per conversation or per matching prompt? | Alerts are created per matching prompt, not just once per conversation. Each matching prompt creates an alert containing the user, prompt text, and time. If the same user submits another prompt that matches later in the same conversation, the article says a separate alert is created for that later prompt as well. | https://kb.expedient.com/docs/ai-ctrl-prompt-monitoring-and-alerting | [
{
"url": "https://kb.expedient.com/docs/ai-ctrl-prompt-monitoring-and-alerting.md",
"excerpt": "Alerts are generated on a per-prompt basis, giving you an accurate, prompt-by-prompt record of matching activity:"
},
{
"url": "https://kb.expedient.com/docs/ai-ctrl-prompt-monitoring-and-alerting.md",
... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0050 | expedient | AI CTRL | how-to | how-to | prompt-alert-query-matching | KB page as retrieved 2026-10-31 | How do I match exact phrases or partial words in a prompt alert? | The monitoring guide uses double quotes for exact phrases and asterisks around a term for partial matches. Matching is not case sensitive. A plain word matches whole words only, so it will not match a longer variant such as “confidentiality”; use the partial-match form when you want variations included. The guide also ... | https://kb.expedient.com/docs/configuring-api-access | [
{
"url": "https://kb.expedient.com/docs/configuring-api-access.md",
"excerpt": "Once access is granted, each user creates their own:"
},
{
"url": "https://kb.expedient.com/docs/configuring-api-access.md",
"excerpt": "The key is shown once. It authenticates as that user and carries that user's mo... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0051 | expedient | AI CTRL | how-to | how-to | compliance-platform-usage-review | KB page as retrieved 2026-10-31 | Where should I look to review AI CTRL prompt and token usage? | The Compliance Guide says gateway activity is stored under the Elastic data view named `prompt`; confirm that view rather than `logs`, which contains unrelated backend system data. Use the AI Gateway Monitoring dashboard for at-a-glance usage and Discover for deeper reports. The guide also recommends excluding embeddin... | https://kb.expedient.com/docs/compliance-guide | [
{
"url": "https://kb.expedient.com/docs/compliance-guide.md",
"excerpt": "All AI gateway activity is stored under the Elastic data view named `prompt`. Always confirm you are on this data view: the default `logs` data view contains unrelated backend system data and will not give you a meaningful picture of ... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0052 | expedient | AI CTRL | faq | reference | prompt-token-growth-explanation | KB page as retrieved 2026-10-31 | Why do prompt token counts grow during a long chat or after attaching files? | Each new message in an ongoing chat carries prior conversation history forward as context, so prompt tokens tend to grow as the conversation continues. Attached files can increase the prompt further because extracted text is injected into it. The Compliance Guide says a prompt-token-to-response-token ratio well above o... | https://kb.expedient.com/docs/compliance-guide | [
{
"url": "https://kb.expedient.com/docs/compliance-guide.md",
"excerpt": "Because each new message in an ongoing chat carries prior conversation history forward as context, prompt tokens tend to grow the longer a conversation continues, and grow further still if data files are attached, since their extracte... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0053 | expedient | AI CTRL | policy | policy | interactive-use-vs-programmatic | KB page as retrieved 2026-10-31 | What does “Unlimited” use mean, and does it cover automated API requests? | The November policy defines Unlimited as reasonable, interactive use by an authenticated individual through the web, desktop, or mobile interface. Requests must be submitted through direct human interaction. API or SDK access, scripts, bots, agents, batch processing, and automated sessions are listed as excluded usage;... | https://kb.expedient.com/docs/ai-ctrl-fair-use-policy-11-1-2026 | [
{
"url": "https://kb.expedient.com/docs/ai-ctrl-fair-use-policy-11-1-2026.md",
"excerpt": "\"Unlimited\" means reasonable, interactive use of AI CTRL by an authenticated individual User through the AI CTRL web, desktop, or mobile interface."
},
{
"url": "https://kb.expedient.com/docs/ai-ctrl-fair-us... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0054 | expedient | AI CTRL | policy | policy | fair-use-excess-usage-response | KB page as retrieved 2026-10-31 | What can Expedient do if a workload creates a sustained impact on shared infrastructure? | The fair-use policy says high usage alone is not a violation; it considers sustained, disproportionate workloads and their impact or reasonable risk to shared infrastructure. When practicable, Expedient will notify the customer and work toward a resolution, which may include modifying the workload, applying technical l... | https://kb.expedient.com/docs/ai-ctrl-fair-use-policy-11-1-2026 | [
{
"url": "https://kb.expedient.com/docs/ai-ctrl-fair-use-policy-11-1-2026.md",
"excerpt": "High usage alone shall not serve as the basis for violation of this Policy."
},
{
"url": "https://kb.expedient.com/docs/ai-ctrl-fair-use-policy-11-1-2026.md",
"excerpt": "Users may not use AI CTRL in a man... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0055 | expedient | AI CTRL | policy | policy | human-review-and-output-responsibility | KB page as retrieved 2026-10-31 | Who is responsible for checking AI CTRL output before using it in a decision? | Users are responsible for reviewing and verifying AI CTRL output because it may be inaccurate, incomplete, misleading, or outdated and may include incorrect citations or other errors. The policy also calls for appropriate human review and professional judgment when output informs decisions that may materially affect so... | https://kb.expedient.com/docs/ai-ctrl-fair-use-policy-11-1-2026 | [
{
"url": "https://kb.expedient.com/docs/ai-ctrl-fair-use-policy-11-1-2026.md",
"excerpt": "AI-generated outputs may be inaccurate, incomplete, misleading, or outdated and may contain incorrect citations, references, or other errors. Users are responsible for reviewing and verifying AI CTRL outputs before re... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0056 | expedient | AI CTRL | faq | reference | universal-credit-ratecard-structure | KB page as retrieved 2026-10-31 | How does the AI CTRL ratecard compare credit use across different models? | The NextGen ratecard uses Universal Credits across models, with token allocations varying by model capability. It sets GPT-6 Luna as the baseline standard model and lists input and output tokens per 1,000 credits for each model. The page notes that higher-capability models consume more credits per token, actual usage v... | https://kb.expedient.com/docs/ai-ctrl-ratecard-nextgen | [
{
"url": "https://kb.expedient.com/docs/ai-ctrl-ratecard-nextgen.md",
"excerpt": "- Higher-capability models consume more credits per token\n- Efficient models stretch your credits further\n- All values based on 1000 Credits\n- **GPT-6 Luna** is the baseline standard model\n- Green values indicate more toke... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0057 | expedient | AI CTRL | service-overview | reference | standard-rag-connector-sources | KB page as retrieved 2026-10-31 | Which standard data sources can Expedient connect for AI CTRL search? | The Standard Data Connectors page lists SharePoint Online, OneDrive Enterprise, Google Drive, Dropbox, Box, Confluence Cloud, ServiceNow, Notion, and Amazon S3. A connector indexes documents so AI CTRL can search by meaning and answer with citations to original files. The article positions this for document lookup, not... | https://kb.expedient.com/docs/ai-ctrl-standard-data-connectors-for-rag | [
{
"url": "https://kb.expedient.com/docs/ai-ctrl-standard-data-connectors-for-rag.md",
"excerpt": "| SharePoint (Cloud / Online) | Microsoft |\n| OneDrive (Enterprise) | Microsoft |\n| Google Drive | Google |\n| Dropbox | Dropbox, Inc. |\n| Box | Box, Inc. |\n| Confluence (Cloud) | Atlassian |\n| ServiceNow ... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0058 | expedient | AI CTRL | how-to | faq | connector-customer-prerequisites | KB page as retrieved 2026-10-31 | What does our organization need to provide before a standard data connector can be configured? | Expedient configures and maintains the connector, while the customer organization creates and maintains the access it uses. The page calls for a dedicated application or service account, read-level permissions to the selected repositories, and credentials delivered securely. The customer also maintains access over time... | https://kb.expedient.com/docs/ai-ctrl-standard-data-connectors-for-rag | [
{
"url": "https://kb.expedient.com/docs/ai-ctrl-standard-data-connectors-for-rag.md",
"excerpt": "Expedient configures and maintains the connector. Your organization creates and maintains the access it uses."
},
{
"url": "https://kb.expedient.com/docs/ai-ctrl-standard-data-connectors-for-rag.md",
... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0059 | expedient | AI CTRL | faq | reference | connector-ingestion-timeline | KB page as retrieved 2026-10-31 | When does a standard connector sync content, and how long does setup usually take? | For a standard integration of 50,000 documents or fewer, the guide estimates five business days after valid credentials are confirmed. The first live sync collects and indexes all available content and may take hours to days depending on volume. After that, the connector runs an incremental sync once daily for changed,... | https://kb.expedient.com/docs/ai-ctrl-standard-data-connectors-for-rag | [
{
"url": "https://kb.expedient.com/docs/ai-ctrl-standard-data-connectors-for-rag.md",
"excerpt": "For a standard integration of 50,000 documents or fewer, expect **5 business days from the point valid credentials are confirmed** — not from the date of the request."
},
{
"url": "https://kb.expedient.... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0060 | expedient | AI CTRL | faq | faq | support-routing-and-urgent-outages | KB page as retrieved 2026-10-31 | Should I open a ticket or contact my AI Client Success Manager for an AI CTRL issue? | Use a support ticket for outages, access problems, reproducible errors, custom model or compliance configuration help, and other work Expedient needs to perform. Training, adoption, credit or plan questions, and feature requests belong with the AI Client Success Manager. For an urgent outage, create a ticket and call t... | https://kb.expedient.com/docs/getting-support-for-ai-ctrl | [
{
"url": "https://kb.expedient.com/docs/getting-support-for-ai-ctrl.md",
"excerpt": "Anything that needs work from us starts as a support ticket; we route it internally from there."
},
{
"url": "https://kb.expedient.com/docs/getting-support-for-ai-ctrl.md",
"excerpt": "Training, enablement, or r... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0101 | expedient | Enterprise Cloud | service-overview | reference | eec-managed-service | KB page as retrieved 2026-10-31 | What does Expedient manage in Enterprise Cloud, and what can I still do? | Expedient Enterprise Cloud is a VMware-powered cloud for running VMware-based workloads on enterprise-class hardware without refactoring applications when moving to a cloud operating model. Expedient manages the hardware platform, software updates, licensing, and ongoing maintenance. Clients retain full administrative ... | https://kb.expedient.com/docs/enterprise-cloud | [
{
"url": "https://kb.expedient.com/docs/enterprise-cloud.md",
"excerpt": "Expedient Enterprise Cloud is a VMware-powered cloud that allows clients to run their VMware-based workloads seamlessly on enterprise-class hardware and avoid refactoring applications while moving into a cloud operating model. Expedie... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0102 | expedient | Enterprise Cloud | service-overview | reference | eec-platform-capabilities | KB page as retrieved 2026-10-31 | What management features and storage choices come with Enterprise Cloud? | Enterprise Cloud supports Windows and Linux, provides a self-service portal, and uses software-defined networking so clients can allocate preferred private IP networks. Dedicated hosts are available for licensing or compliance needs. The page lists REST API and CLI management access and two-factor authentication, with ... | https://kb.expedient.com/docs/enterprise-cloud | [
{
"url": "https://kb.expedient.com/docs/enterprise-cloud.md",
"excerpt": "- Pay per GB of RAM\n- Pay per GB of storage allocated\n- Hyperconverged, all-flash storage, encrypted at rest\n - Non-hyper-converged storage is available\n- All-flash storage\n- Support for Windows and Linux operating systems\n- So... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0103 | expedient | Enterprise Cloud | policy | reference | eec-deployment-baseline | KB page as retrieved 2026-10-31 | What is configured by default when Enterprise Cloud is deployed? | The documented deployment sets up client private networking on the Expedient platform, tenant resource pools, and external tenant networking through a firewall. Clients are given self-service portal access, and their user directory, such as Active Directory, is configured for authentication. Optional migration services... | https://kb.expedient.com/docs/enterprise-cloud | [
{
"url": "https://kb.expedient.com/docs/enterprise-cloud.md",
"excerpt": "- Client private networking will be configured on the Expedient platform\n- Tenant resource pools will be configured\n- External networking will be configured for client tenants through a firewall\n- Clients will be given access and a... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0104 | expedient | Private Cloud | service-overview | reference | private-cloud-overview | KB page as retrieved 2026-10-31 | What does Expedient Private Cloud provide, and what remains under my control? | Private Cloud is managed infrastructure-as-a-service for VMware- or Nutanix-based workloads on enterprise-class hardware. Expedient maintains the hardware platform, software updates, licensing, and ongoing maintenance, while clients retain full administrative access to their VMs and many of the operations possible in t... | https://kb.expedient.com/docs/private-cloud | [
{
"url": "https://kb.expedient.com/docs/private-cloud.md",
"excerpt": "Private Cloud is a fully managed infrastructure-as-a-service that allows clients to run their VMware or Nutanix based workloads seamlessly on enterprise-class hardware and avoid refactoring applications while moving into a cloud operatin... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0105 | expedient | Private Cloud | service-overview | reference | private-cloud-hypervisors | KB page as retrieved 2026-10-31 | Which hypervisors and management interfaces does Private Cloud offer? | Private Cloud offers VMware Cloud Foundation or Nutanix Cloud Infrastructure. The page describes a single management pane linked across sites, with VMware vCenter per site or Nutanix Prism Central per site. Listed features also include REST API and CLI management access and two-factor authentication. Hosts are dedicate... | https://kb.expedient.com/docs/private-cloud | [
{
"url": "https://kb.expedient.com/docs/private-cloud.md",
"excerpt": "- Choice of hypervisor\n - VMware Cloud Foundation\n - Nutanix Cloud Infrastructure\n- Pay per core\n- Pay per GB of storage allocated\n- All-flash storage, encrypted at rest\n - VMware: Standard and Performance Storage\n - Nutanix: ... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0106 | expedient | Private Cloud | policy | reference | private-cloud-kms | KB page as retrieved 2026-10-31 | How are Private Cloud networking and key management set up? | The default deployment includes client private networking on Expedient’s platform, dedicated hosts, external tenant networking through a firewall, and access to the management plane. A client directory such as Active Directory is configured for authentication. Native KMS is standard for Nutanix Private Cloud; external ... | https://kb.expedient.com/docs/private-cloud | [
{
"url": "https://kb.expedient.com/docs/private-cloud.md",
"excerpt": "- Client private networking will be configured on the Expedient platform\n- Hosts are dedicated to each client\n- External networking will be configured for client tenants through a firewall\n- Clients will be given access and authentica... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0107 | expedient | Data Protection | service-overview | reference | cloud-data-protection-overview | KB page as retrieved 2026-10-31 | What is Expedient Cloud Data Protection? | Cloud Data Protection is a Backup as a Service offering. Expedient manages the platform and capacity so clients can focus on the data and applications they protect. Clients can recover data on their schedule, and Expedient can assist with recoveries. The page describes support for multiple operating systems, databases,... | https://kb.expedient.com/docs/data-protection | [
{
"url": "https://kb.expedient.com/docs/data-protection.md",
"excerpt": "Cloud Data Protection is Backup as a Service that allows clients to protect their data while removing the overhead of maintaining capacity and management of the platform itself. As a managed service, Expedient maintains the platform an... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0108 | expedient | Data Protection | service-overview | reference | cloud-data-protection-features | KB page as retrieved 2026-10-31 | What protection and service features are included in Cloud Data Protection? | Listed features include encryption at rest, immutable backups, deduplication and compression, customizable retention, and archiving to object storage. The service is available on multi-tenant or dedicated platforms and includes a management portal and 24x7x365 backup and restore assistance. Replication to Expedient and... | https://kb.expedient.com/docs/data-protection | [
{
"url": "https://kb.expedient.com/docs/data-protection.md",
"excerpt": "- Pay per TB used\n- Encryption at rest\n- Immutable backups\n- Deduplication and compression\n- Available as a multi-tenant or dedicated platform\n- Easy to use management portal\n- 24x7x365 support for backup and restore assistance\n... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0109 | expedient | Data Protection | policy | reference | cloud-data-protection-defaults | KB page as retrieved 2026-10-31 | What are the default Cloud Data Protection retention periods? | Default retention is two or four weeks, with custom retention periods available. Enterprise Cloud VMs are backed up from creation. The page also lists deduplication, compression, and erasure coding as enabled, along with encryption at rest and in flight. These are the default deployment settings described in the Data P... | https://kb.expedient.com/docs/data-protection | [
{
"url": "https://kb.expedient.com/docs/data-protection.md",
"excerpt": "- 2 week or 4 week retention\n - Custom retention periods available\n- All Enterprise Cloud VMs backed up from creation\n- Deduplication, compression and erasure coding enabled\n- Encryption-at-rest and in-flight enabled"
}
] | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0110 | expedient | Disaster Recovery | service-overview | reference | draaS-service | KB page as retrieved 2026-10-31 | What is Expedient Push Button DR and who can initiate a failover? | Push Button DR is Expedient’s Disaster Recovery as a Service product for protecting applications and recovering them after a disaster. Expedient manages the platform and recovery resources. Clients can declare disasters and initiate failovers, and Expedient is also available to initiate failover and assist with recover... | https://kb.expedient.com/docs/disaster-recovery | [
{
"url": "https://kb.expedient.com/docs/disaster-recovery.md",
"excerpt": "Expedient's \"Push Button DR\" is a Disaster Recovery as a Service (DRaaS) product that allows you to protect your applications and recover quickly in the event of a disaster. As a managed service, Expedient maintains the platform an... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0111 | expedient | Disaster Recovery | policy | reference | draaS-test-bubble | KB page as retrieved 2026-10-31 | How does the DR service isolate tests from production? | The default DR deployment provides a “test bubble” network intended to prevent IP conflicts with production during testing. The page says the bubble also prevents application, data, and IT service conflicts while validating the test. Clients can contact the Operations Support Center to initiate a disaster failover or s... | https://kb.expedient.com/docs/disaster-recovery | [
{
"url": "https://kb.expedient.com/docs/disaster-recovery.md",
"excerpt": "- Client networking will be configured on the Expedient platform\n- Workload protection will be configured with the Client's input\n- Networking will be configured for failover between Expedient locations\n- Client will be provided a... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0112 | expedient | Disaster Recovery | policy | reference | draaS-post-failover | KB page as retrieved 2026-10-31 | Are there resource charges or follow-up requirements after a DR failover? | Push Button DR On Demand subscribers are charged for resource consumption during tests and failovers. After clients declare a disaster or commit to failover, they are asked to reside at the secondary site while reverse replication completes, typically 24–48 hours depending on environment size and change rate. Data prot... | https://kb.expedient.com/docs/disaster-recovery | [
{
"url": "https://kb.expedient.com/docs/disaster-recovery.md",
"excerpt": "- Client networking will be configured on the Expedient platform\n- Workload protection will be configured with the Client's input\n- Networking will be configured for failover between Expedient locations\n- Client will be provided a... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0113 | expedient | Disaster Recovery | faq | reference | draaS-test-pricing | KB page as retrieved 2026-10-31 | What are the DR test charges for On Demand versus traditional Push Button DR? | Push Button DR on Demand subscribers are billed for RAM, compute, and VMware licensing based on utilization for each day VMs are failed over for testing, but Expedient does not charge labor to conduct the test. Traditional Push Button DR subscribers do not incur additional expenses for properly scheduled DR tests. Thes... | https://kb.expedient.com/docs/disaster-recovery-tests | [
{
"url": "https://kb.expedient.com/docs/disaster-recovery-tests.md",
"excerpt": "Clients who subscribe to **Push Button DR on Demand** will be billed for RAM, Compute, and VMware licensing based on utilization each day any VMs are failed over for testing. Expedient does not charge for labor to conduct the t... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0114 | expedient | Disaster Recovery | how-to | support-intake-scenario | draaS-request-test | KB page as retrieved 2026-10-31 | How do I request a DR test after implementation? | After the initial implementation test, request another test through Expedient’s Operations Support Center. Tests must be scheduled several weeks ahead, and OSC will try to accommodate requested dates subject to staff capacity and existing bookings. For the first test, Service Delivery and the project manager conduct it... | https://kb.expedient.com/docs/disaster-recovery-tests | [
{
"url": "https://kb.expedient.com/docs/disaster-recovery-tests.md",
"excerpt": "After your initial implementation, you may request to schedule an additional DR test with Expedient's Operations Support Center (OSC). Please note that these must be scheduled several weeks in advance. OSC staff will attempt to... | first-party-draft | true | expedient-first-party-content | documented-procedure | expedient-kb | null | null | null | null |
expkb-0115 | expedient | Disaster Recovery | service-overview | reference | draaS-bubble-test | KB page as retrieved 2026-10-31 | What happens during a DR bubble test? | A Bubble Test powers on replicated server copies at the recovery site in an isolated network while production remains up. It allows validation of server OS and data integrity, but full application testing may not be available. Bubble servers have no inter-VLAN or internet connectivity and can communicate only with serv... | https://kb.expedient.com/docs/disaster-recovery-tests | [
{
"url": "https://kb.expedient.com/docs/disaster-recovery-tests.md",
"excerpt": "Depending on what test is performed, the impact on your environment will differ. There are generally three types of tests: * **Bubble Test:** During a Bubble test, a replicated copy of your servers are powered on at the recover... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0116 | expedient | Disaster Recovery | service-overview | reference | draaS-live-failover-impact | KB page as retrieved 2026-10-31 | How much downtime should we expect while a live failover test runs? | During a Live Failover test, production servers are powered off while replicated copies are built and powered on at the recovery site. The environment is effectively down during the network and server failover. Expedient usually allocates about an hour, although timing varies with replication-group size and how quickly... | https://kb.expedient.com/docs/disaster-recovery-tests | [
{
"url": "https://kb.expedient.com/docs/disaster-recovery-tests.md",
"excerpt": "Depending on what test is performed, the impact on your environment will differ. There are generally three types of tests: * **Bubble Test:** During a Bubble test, a replicated copy of your servers are powered on at the recover... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0117 | expedient | Disaster Recovery | policy | reference | draaS-no-commit-changes | KB page as retrieved 2026-10-31 | What happens to the recovery-site servers in a no-commit test? | In a Live Failover test with no commit, the failover is canceled after network and server groups are failed over and validated. Servers at the recovery site are powered off and deleted, and production servers are powered back on in the same state as when the test began. Changes made to recovery-site servers are lost wh... | https://kb.expedient.com/docs/disaster-recovery-tests | [
{
"url": "https://kb.expedient.com/docs/disaster-recovery-tests.md",
"excerpt": "Depending on what test is performed, the impact on your environment will differ. There are generally three types of tests: * **Bubble Test:** During a Bubble test, a replicated copy of your servers are powered on at the recover... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0118 | expedient | Disaster Recovery | service-overview | reference | nutanix-recovery-plan-purpose | KB page as retrieved 2026-10-31 | What does a Nutanix recovery plan orchestrate? | A recovery plan orchestrates restoration of protected VMs at a backup location. It can recover all specified VMs at once or use power-on sequences with optional interstage delays to restore dependent applications in order. Plans can run embedded custom scripts when NGT is installed and provide options to change or main... | https://kb.expedient.com/docs/creating-recovery-plans | [
{
"url": "https://kb.expedient.com/docs/creating-recovery-plans.md",
"excerpt": "A recovery plan orchestrates restoring protected VMs at a backup location. Recovery plans can either recover all specified VMs at once or, with what is essentially a runbook functionality, use power-on sequences with optionally... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0119 | expedient | Disaster Recovery | how-to | how-to | nutanix-create-recovery-plan | KB page as retrieved 2026-10-31 | How do I create a Nutanix recovery plan in Prism Central? | In Prism Central, open Data Protection > Recovery Plans, select Create New Recovery Plan, name the policy, and add primary and recovery locations. Add VMs, or change the dropdown from VM name to Category. If multiple categories are created, recovery stages can be configured; otherwise the page says to select Environmen... | https://kb.expedient.com/docs/creating-recovery-plans | [
{
"url": "https://kb.expedient.com/docs/creating-recovery-plans.md",
"excerpt": "- Data Protection > Recovery Plans\n - Create New Recovery Plan\n - Name Policy\n- Select locations:"
},
{
"url": "https://kb.expedient.com/docs/creating-recovery-plans.md",
"excerpt": "Add VMs"
}
] | first-party-draft | true | expedient-first-party-content | documented-procedure | expedient-kb | null | null | null | null |
expkb-0120 | expedient | Disaster Recovery | how-to | how-to | nutanix-recovery-plan-stages | KB page as retrieved 2026-10-31 | How can a Nutanix recovery plan sequence dependent applications? | Configure multiple recovery stages and add delays between them for applications with dependencies. The article also says you may enable scripts for a selected VM; that requires NGT and placing scripts in the specified Windows or Linux directories. Use the environment-specific Production or Test path listed in the artic... | https://kb.expedient.com/docs/creating-recovery-plans | [
{
"url": "https://kb.expedient.com/docs/creating-recovery-plans.md",
"excerpt": "- Multiple stages can be created with delay’s in between if restoring specific applications that have dependencies.\n- By selection a VM you my also enable scripts to be ran. This requires NGT to be installed and the scripts to... | first-party-draft | true | expedient-first-party-content | documented-procedure | expedient-kb | null | null | null | null |
expkb-0121 | expedient | Disaster Recovery | policy | reference | nutanix-network-prerequisites | KB page as retrieved 2026-10-31 | What network mappings and Guest Tools are needed for DR? | Network mappings are configured for planned, unplanned, and test failovers, with networks designated under their respective sites. Select the appropriate network; if using static IPs, custom IP mapping can be configured in advanced settings. The guide warns that Nutanix Guest Tools must be installed on every VM selecte... | https://kb.expedient.com/docs/creating-recovery-plans | [
{
"url": "https://kb.expedient.com/docs/creating-recovery-plans.md",
"excerpt": "Network mappings are configured here for the planned/unplanned failovers and test failover. The networks are designated as such under their respective sites."
},
{
"url": "https://kb.expedient.com/docs/creating-recovery... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0122 | expedient | Disaster Recovery | policy | planning-scenario | nutanix-rpo-options | KB page as retrieved 2026-10-31 | What RPO settings does Expedient recommend for Nutanix DR? | The best-practices page recommends Async replication with a one-hour RPO. NearSync may be configured for an RPO from one to 15 minutes, but lower RPOs are more sensitive to bandwidth limitations or high change rates. Synchronous replication is not supported because it requires extremely stringent latency. | https://kb.expedient.com/docs/disaster-recovery-best-practices | [
{
"url": "https://kb.expedient.com/docs/disaster-recovery-best-practices.md",
"excerpt": "**RPO Setting**"
}
] | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0123 | expedient | Disaster Recovery | policy | planning-scenario | nutanix-separate-plans | KB page as retrieved 2026-10-31 | Should Async and NearSync VMs use separate recovery plans? | The guide says to configure Async and NearSync replicated VMs on separate plans. It also recommends non-routable networks for testing to avoid interfering with production. For an Expedient recovery target, it gives a test bubble network with a non-routable network or subnet such as APIPA 169.254.0.0 as an example. | https://kb.expedient.com/docs/disaster-recovery-best-practices | [
{
"url": "https://kb.expedient.com/docs/disaster-recovery-best-practices.md",
"excerpt": "- Always configure Async and NearSync replicated VMs on separate plans.\n- Implement non-routable networks for testing to prevent interference with production environments.\n - If Expedient is the recovery target for ... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0124 | expedient | Disaster Recovery | policy | reference | nutanix-plan-validation-cleanup | KB page as retrieved 2026-10-31 | How should I validate and clean up recovery plan tests? | Run the Validate workflow after making recovery-plan changes. For trials, use the Test and Clean-Up workflows to manage VM states. After running Test, run Clean-Up instead of manually deleting VMs. These recommendations describe how to manage VM state during recovery-plan trials. | https://kb.expedient.com/docs/disaster-recovery-best-practices | [
{
"url": "https://kb.expedient.com/docs/disaster-recovery-best-practices.md",
"excerpt": "> [!WARNING]\n> NGT Installation\n> \n> NGT is strongly recommended to avoid the need for later installation across many servers and to mitigate networking issues during failover."
}
] | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0125 | expedient | Disaster Recovery | policy | reference | nutanix-third-party-backups | KB page as retrieved 2026-10-31 | How should I handle Cohesity or Commvault backups during failover? | If using Cohesity or Commvault, the guide recommends disabling backups before a DR failover to prevent snapshot-induced data corruption during transition. If the primary site is offline after failover, it recommends creating a new protection policy to take local snapshots or recovery points for production VMs. Addition... | https://kb.expedient.com/docs/disaster-recovery-best-practices | [
{
"url": "https://kb.expedient.com/docs/disaster-recovery-best-practices.md",
"excerpt": "If using backup solutions like Cohesity or Commvault, disable backups before a DR failover. This preemptive step prevents snapshot-induced data corruption during the transition."
},
{
"url": "https://kb.expedie... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0126 | expedient | Data Protection | faq | reference | cohesity-storage-domain | KB page as retrieved 2026-10-31 | What is a Cohesity storage domain? | A Cohesity storage domain is a logical object containing all data written from backups. It includes physical backup data and overhead such as data resiliency, an index of backup jobs, and garbage collection. This helps explain why reported domain utilization can include more than the raw backup data itself. | https://kb.expedient.com/docs/cloud-data-protection-faq | [
{
"url": "https://kb.expedient.com/docs/cloud-data-protection-faq.md",
"excerpt": "A storage domain is a logical object that contains all of the data written from backups. This includes the physical data from the backups themselves, but also overhead such as data resiliency, an index of backup jobs, and gar... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0127 | expedient | Data Protection | faq | reference | cohesity-protection-job | KB page as retrieved 2026-10-31 | What is a protection job in Cohesity? | A protection job is a logical object that defines what needs to be backed up, how it should be backed up, and which retention policy should maintain it. The FAQ also clarifies that a protection job snapshot is a synonymous term for a backup job. | https://kb.expedient.com/docs/cloud-data-protection-faq | [
{
"url": "https://kb.expedient.com/docs/cloud-data-protection-faq.md",
"excerpt": "A protection job is a logical object that defines what needs to be backed up, how it should be backed up, and what retention policy should be used to maintain those protection jobs."
},
{
"url": "https://kb.expedient.... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0128 | expedient | Data Protection | faq | reference | cohesity-storage-metrics | KB page as retrieved 2026-10-31 | How should I read Cohesity storage-domain reporting metrics? | Logical Data is hydrated or front-end data that would have been backed up without compression and deduplication. Data In is the estimated amount sent before those processes, while Data Written is utilization after them. Resiliency Impact is overhead protecting written data, and Storage Consumed Local and Cloud adds Dat... | https://kb.expedient.com/docs/cloud-data-protection-faq | [
{
"url": "https://kb.expedient.com/docs/cloud-data-protection-faq.md",
"excerpt": "| **Logical Data** | Also known as front-end data or hydrated data. This number represents the total amount of data that was backed up if compression and deduplication did not take place. |\n| --- | --- |\n| **Data In** | Thi... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0129 | expedient | Data Protection | faq | reference | cohesity-erasure-overhead | KB page as retrieved 2026-10-31 | Why is Resiliency Impact around 50% of Data Written? | The FAQ says this is expected overhead because Expedient standardizes a 4+2 Erasure Coding scheme. The article describes Resiliency Impact as utilization overhead intended to protect Data Written in the event of hardware failure, and says the normal resiliency footprint is 50% of Data Written. | https://kb.expedient.com/docs/cloud-data-protection-faq | [
{
"url": "https://kb.expedient.com/docs/cloud-data-protection-faq.md",
"excerpt": "This is the expected overhead, as Expedient standardizes a 4+2 Erasure Coding scheme."
},
{
"url": "https://kb.expedient.com/docs/cloud-data-protection-faq.md",
"excerpt": "| **Logical Data** | Also known as front... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0130 | expedient | Data Protection | troubleshooting | reference | cohesity-reclamation-delay | KB page as retrieved 2026-10-31 | I deleted backup snapshots; why has storage utilization not dropped? | Deleting a snapshot does not immediately reclaim storage, even if deletion happened before the retention date. Cohesity performs internal operations to validate and earmark data before deletion. The FAQ says this can take 5–7 days before space is largely reclaimed, so an immediate storage quota decrease is not expected... | https://kb.expedient.com/docs/cloud-data-protection-faq | [
{
"url": "https://kb.expedient.com/docs/cloud-data-protection-faq.md",
"excerpt": "Space reclamation is not immediate despite the snapshot being deleted prior to a retention date. This is to be expected, as Cohesity runs a series of complex internal operations to validate and earmark blobs of data before de... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0131 | expedient | Data Protection | policy | reference | ahv-volume-group-support | KB page as retrieved 2026-10-31 | Can Cohesity protect Nutanix AHV Volume Groups? | The caveats page says Cohesity does not currently support backup or restoration of Volume Groups and does not support Instant Volume Mounts on AHV clusters. Expedient recommends using data protection services built into AOS for Volume Groups. For individual-file restoration, the page says a Cohesity agent must be insta... | https://kb.expedient.com/docs/cloud-data-protection-caveats | [
{
"url": "https://kb.expedient.com/docs/cloud-data-protection-caveats.md",
"excerpt": "# Cloud Data Protection Caveats"
}
] | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0132 | expedient | Data Protection | policy | reference | cloud-protection-roles | KB page as retrieved 2026-10-31 | Who manages retention policies, backup jobs, and restores? | The responsibility matrix marks retention policies, backup jobs, and replication jobs as co-managed. Expedient creates default two- and four-week policies, while clients may create additional policies to meet retention needs. VMware and Windows restores are also co-managed; the matrix notes recovery may be initiated by... | https://kb.expedient.com/docs/data-protection | [
{
"url": "https://kb.expedient.com/docs/data-protection.md",
"excerpt": "| **Cloud Data Protection Responsibility Matrix** | | | | |\n| --- | --- | --- | --- | --- |\n| **Task** | **Expedient** | **Client** | **Co-Managed** | **Co-Managed tasks can be performed by Expedient or Client based on Client's p... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0133 | expedient | Data Protection | how-to | how-to | cohesity-vm-restore | KB page as retrieved 2026-10-31 | How do I recover an entire virtual machine from Cohesity backups? | Open Data Protection > Recoveries and choose Recover > VMs. Search by VM name or protection policy; wildcards and partial names are supported. Select the VM and configure task name, recovery point, and location. The article strongly recommends using Detach Network and adding networking in Enterprise Cloud after restora... | https://kb.expedient.com/docs/how-to-restore-a-virtual-machine | [
{
"url": "https://kb.expedient.com/docs/how-to-restore-a-virtual-machine.md",
"excerpt": "1. [Log in](/v1/docs/how-to-access-cloud-data-protection) to the Cloud Data Protection system.\n2. From the navigation bar at the top of the page, select the **Data Protection** → **Recoveries** tab. \n![](https://cdn... | first-party-draft | true | expedient-first-party-content | documented-procedure | expedient-kb | null | null | null | null |
expkb-0134 | expedient | Data Protection | how-to | how-to | cohesity-volume-mount | KB page as retrieved 2026-10-31 | How do I mount a restored backup volume to a Windows VM and copy data? | In Cloud Data Protection, go to Data Protection > Recoveries and choose Recover > Instant Volume Mount. Find the source server, select the volume and recovery point, choose the target VM, and create the task. After the VMDK attaches, use Windows Disk Management to bring the disk online and copy data, then take it offli... | https://kb.expedient.com/docs/how-to-restore-a-volume-to-a-windows-virtual-machine | [
{
"url": "https://kb.expedient.com/docs/how-to-restore-a-volume-to-a-windows-virtual-machine.md",
"excerpt": "1. [Login](/v1/docs/how-to-access-cloud-data-protection) to the respective Cloud Data Protection system\n2. From the navigation bar at the top of the page, select the **Data Protection** → **Recover... | first-party-draft | true | expedient-first-party-content | documented-procedure | expedient-kb | null | null | null | null |
expkb-0135 | expedient | Data Protection | how-to | how-to | cohesity-sql-restore | KB page as retrieved 2026-10-31 | What steps restore an MS SQL database from Cohesity? | Open Data Protection Recoveries, choose Recover > MS SQL, and search for the SQL server or database; wildcards and partial names are supported. Select the database and snapshot, configure Recovery Options, and select Recover. For overwriting an Always On Availability Group database, additional actions are required, inc... | https://kb.expedient.com/docs/how-to-restore-a-sql-database-to-a-windows-virtual-machine | [
{
"url": "https://kb.expedient.com/docs/how-to-restore-a-sql-database-to-a-windows-virtual-machine.md",
"excerpt": "1. [Log in](/v1/docs/how-to-access-cloud-data-protection) to the Cloud Data Protection system.\n2. Select the Data Protection Recoveries tab from the navigation bar at the top of the page. \n... | first-party-draft | true | expedient-first-party-content | documented-procedure | expedient-kb | null | null | null | null |
expkb-0136 | expedient | Data Protection | how-to | how-to | cohesity-linux-file-restore | KB page as retrieved 2026-10-31 | How do I restore individual files from a Linux VM backup? | Ensure the Linux VM you will recover to has prerequisites from the helper-app article. In Cloud Data Protection, open Recoveries and select Recover > Files or Folders. Search by file name or path; wildcards and partial names are supported. Select the item and snapshot, choose Recover to Server, configure recovery optio... | https://kb.expedient.com/docs/how-to-restore-files-and-folders-from-a-linux-virtual-machine | [
{
"url": "https://kb.expedient.com/docs/how-to-restore-files-and-folders-from-a-linux-virtual-machine.md",
"excerpt": "1. Ensure the Linux VM you wish to recover to has the proper prerequisites in place as detailed on the [How-to Install the Cohesity Helper App (Linux)](/v1/docs/how-to-install-the-cohesity-... | first-party-draft | true | expedient-first-party-content | documented-procedure | expedient-kb | null | null | null | null |
expkb-0137 | expedient | Colocation | service-overview | reference | colocation-service-overview | KB page as retrieved 2026-10-31 | What does Expedient Colocation provide, and who manages facility operations? | Expedient Colocation lets clients host physical hardware in hardened, resilient data centers staffed around the clock across the United States. Expedient manages the power distribution, cooling, rack placement, and floor space. Clients retain full access to their hardware and can submit tickets to grant access to vendo... | https://kb.expedient.com/docs/colocation | [
{
"url": "https://kb.expedient.com/docs/colocation.md",
"excerpt": "Expedient Colocation enables clients to host their physical hardware in hardened, resilient data centers staffed 24x7x365 across the United States. Expedient manages the power distribution, cooling, rack placement and floor space in each of... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0138 | expedient | Colocation | service-overview | reference | colocation-facility-resilience | KB page as retrieved 2026-10-31 | What power, environmental, and connectivity features does Colocation list? | The Colocation page lists multiple power feeds, A+B circuits to the rack, multiple public utility feeds, redundant UPS, and redundant generators. It also lists temperature and humidity monitoring, diverse carrier feeds, Expedient service connectivity, line-rate switch ports, low latency, and no egress fees. Facility-sp... | https://kb.expedient.com/docs/colocation | [
{
"url": "https://kb.expedient.com/docs/colocation.md",
"excerpt": "- Metered or circuit based power\n- Multiple rack and cage options\n - Cages\n - Traditional cage with available options:\n - Available options:\n - Roof\n - Under floor\n - Biometric security\n - Slimline C... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0139 | expedient | Edge | service-overview | reference | edge-managed-service | KB page as retrieved 2026-10-31 | What does Expedient Edge manage, and what remains under client control? | Expedient Edge is managed infrastructure for workloads at edge locations on redundant hardware, backed by a highly available cloud management plane. Expedient maintains the platform, including break-fix, software updates, licensing, and ongoing maintenance. Clients have full administrative access to their virtual machi... | https://kb.expedient.com/docs/edge | [
{
"url": "https://kb.expedient.com/docs/edge.md",
"excerpt": "Expedient Edge is a fully managed infrastructure for running workloads at edge locations on redundant hardware backed by a highly available cloud management plane, support, and maintenance from Expedient. As a managed service, Expedient maintains... | first-party-draft | true | expedient-first-party-content | no-system-changes | expedient-kb | null | null | null | null |
expkb-0140 | expedient | Colocation | how-to | support-intake-scenario | remote-hands-request-process | KB page as retrieved 2026-10-31 | How do I submit a remote hands request? | All remote hands requests require a ticket. For standard work, contact the Operations Support Center and provide a detailed description in an SMC ticket. For comprehensive support or specific requirements, use the Project Change Request process by sending a written request to your dedicated account manager. In an emerg... | https://kb.expedient.com/docs/remote-hands-support | [
{
"url": "https://kb.expedient.com/docs/remote-hands-support.md",
"excerpt": "**How to Request Remote Hands Support**"
}
] | first-party-draft | true | expedient-first-party-content | documented-procedure | expedient-kb | null | null | null | null |
End of preview. Expand in Data Studio
Expedient Engram — one consolidated knowledge dataset
1,417 Q&A rows in a single train.jsonl, uniting every Expedient-related dataset
built to date under one schema with a collection tag per row. Built 2026-10-31.
Private repository (contains first-party Expedient content and pending-rights
vendor drafts).
Collections
| collection | rows | contents | rights |
|---|---|---|---|
expedient-kb |
218 | First-party Q&A from Expedient's public customer KB (kb.expedient.com), 126 pages, verbatim-excerpt verified | Expedient first-party content |
expedient-company |
40 | Company, services, data centers (10 markets, 35 MW), compliance (SOC/HIPAA/PCI/GDPR), Operations Support Center, awards, press | Expedient first-party content |
vendor-vmware |
504 | VMware Cloud Foundation 9 architecture, products, licensing, Broadcom support, operations | Third-party-derived; authored summaries (see card) |
vendor-nutanix |
522 | Nutanix platform, NKP, support programs, licensing | Third-party-derived; authored summaries |
vendor-zerto |
71 | Zerto architecture, 10.9/10.9.10 release notes, lifecycle matrix, support | Third-party-derived; authored summaries |
vendor-cohesity |
62 | Cohesity DataProtect, NetBackup 11.2.0.1, FortKnox, 2026 license terms | Third-party-derived; authored summaries |
Schema
Core fields on every row: id (unique across the engram: vmw-, ntx-, zer-,
coh-, expkb-, expco- prefixes), vendor, product, category, question,
answer, source, collection. Vendor-collection rows additionally carry
record_type, intent_family, version_scope, evidence, and draft gates
(review_status, training_eligible, rights_status, execution_level) as
described on each source dataset card. KB and company rows are first-party
(training_eligible: true, rights_status: expedient-first-party-content);
vendor rows remain training_eligible: false / rights_status: pending-review.
Provenance
expedient-kbandexpedient-company: Expedient's own public web properties (kb.expedient.comllms.txtindex; expedient.com service/data-center/support pages), retrieved 2026-10-30/31, with verbatim excerpts retained locally and verified before assembly.- Vendor collections: official Broadcom/VMware, Nutanix, HPE/Zerto and Cohesity pages, with per-row evidence excerpts and retained source text as documented on each source dataset card (vmware-vcf9-knowledge, nutanix-knowledge, zerto-knowledge, cohesity-knowledge).
- IDs are globally unique; no row appears twice.
validation_report.jsonin each source folder documents the per-collection checks.
Intended use
Single-corpus retrieval (RAG) and evaluation for Expedient-facing assistants: the KB and company collections answer "how does Expedient work," and the vendor collections answer "how does the underlying technology work." All rows are question/answer pairs suitable for grounding and eval; vendor rows are additionally gated for training until rights review completes.
Interactive architecture diagram
An interactive HTML diagram visualizing this dataset's knowledge architecture ships with
the repo at diagrams/expedient-company-architecture.html — open it in any browser (self-contained; CDN dependencies:
Tailwind, Lucide icons, Google Fonts). Click any node for the dataset facts behind it
(citations reference dataset row IDs), filter primary/secondary flows, and use the
gear-icon Configure Panel to edit every label in place. Node facts come only from this
dataset's verified rows.
Limitations
- Vendor rows are authored summaries of official pages as retrieved 2026-10-30/31; release status, licensing and compatibility facts are time-sensitive and scoped in-row.
- The KB collection covers 126 of 381 KB pages (v1); long-tail how-tos can be expanded.
- One KB index entry (ai-ctrl.md) was a dead link at retrieval.
License and use — not for resale
This dataset packages publicly available information about how to support and use products: Expedient's own public knowledge base and website, and publicly available vendor product documentation and support resources (VMware/Broadcom, Nutanix, HPE/Zerto, Cohesity), restructured into question/answer form.
- Not for resale. This dataset may be used for research, evaluation, retrieval (RAG), and training personal or internal assistants, but it may not be sold, sublicensed, or offered as a paid data product.
- All product names, trademarks, and upstream documentation remain the property of their respective owners (Expedient, Broadcom/VMware, Nutanix, Hewlett Packard Enterprise/Zerto, Cohesity). Inclusion in this dataset does not transfer any rights.
- Content is provided as-is for informational use; upstream documentation changes over time and rows carry their retrieval dates. Nothing here is a substitute for the vendors' official documentation, contracts, or support agreements.
- Expedient first-party content remains Expedient property.
- Downloads last month
- 93