Spaces:
Running
Running
File size: 3,589 Bytes
f2c0bd9 366223a f2c0bd9 366223a f2c0bd9 366223a f2c0bd9 366223a f2c0bd9 366223a f2c0bd9 366223a f2c0bd9 366223a 187de79 366223a 187de79 366223a 187de79 366223a 67e3963 366223a |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 |
---
title: BizIntel AI Ultra
emoji: 📊
colorFrom: indigo
colorTo: blue
sdk: docker
app_file: app.py
pinned: false
---
# 📊 BizIntel AI Ultra – Advanced Analytics & Agentic BI
**BizIntel AI Ultra** turns any data table—whether a quick CSV upload or a live
database table—into an interactive **insight & forecasting workspace** powered
by:
* **Google Gemini 1.5 Pro** ⇒ natural-language strategy briefs
* **ARIMA time-series pipeline** ⇒ 90-step (or 3-period) forecasts with
model explainability, back-testing, residual ACF & CI bands
* **Plotly Dark™** trend, forecast & EDA visuals
* **Streamlit** front-end running inside a Hugging Face Space
* **Modular tools** (`csv_parser`, `plot_generator`, `forecaster`,
`visuals`, `db_connector`) for quick reuse in your own projects
---
## 🚀 Live Demo (Hugging Face Space)
<https://huggingface.co/spaces/your-handle/BizIntel_AI>
Upload a CSV or Excel (≤ 500 MB) **or** connect to PostgreSQL / MySQL /
SQL Server / Oracle via a SQLAlchemy URI, select a date column and a numeric
metric, and watch:
1. Trend chart
2. 90-step forecast with confidence fan, summary stats & interpret-ability
3. Gemini-generated strategy report
4. Optional histograms, scatter matrices, correlation heat-maps
---
## ✨ Features
| Area | Details |
| ---- | ------- |
| **Data ingest** | CSV, XLS/XLSX, or DB fetch (`sqlalchemy` URI) |
| **Sampling** | Auto-samples giant CSVs (≥ 1 M rows) to avoid OOM |
| **Quick scan** | Markdown dataset overview: schema, missing %, numeric `describe()` |
| **Trend** | Interactive line + markers (Plotly Dark) |
| **Forecast** | ARIMA(1,1,1) auto-freq, 90-step (daily) or 3-step (monthly, etc) |
| | CI fan, residual ACF plot, 20 % back-test (MAPE, RMSE) |
| | Seasonal decomposition (toggle) |
| **Explainability** | Model summary table, plain-English coef notes |
| **Gemini 1.5 Pro** | 5 insights, 3 strategies, risks, extra-visuals → Markdown + download |
| **EDA** | Histograms (with KDE), box / violin plots, scatter-matrix, corr heat-map |
| **Static images** | All figures also saved to `/tmp/*.png` for downstream reports |
| **Extensible** | Each visual is a standalone helper in `tools/` |
---
## 🛠 Local Dev Setup
```bash
# 1. Clone
git clone https://github.com/your-handle/BizIntel_AI.git
cd BizIntel_AI
# 2. Python env
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
# 3. Gemini key
export GEMINI_APIKEY="your_google_ai_studio_key"
# 4. Run
streamlit run app.py
Note : ARIMA & seasonal decomposition come from
statsmodels; Plotly PNG export needs kaleido.
These are included in requirements.txt.
🔌 Database Connect String Examples
text
Copy code
# PostgreSQL
postgresql+psycopg2://user:password@host:port/dbname
# MySQL
mysql+pymysql://user:password@host/dbname
# SQL Server
mssql+pyodbc://user:password@host:1433/db?driver=ODBC+Driver+17+for+SQL+Server
# Oracle
oracle+cx_oracle://user:password@host:1521/?service_name=ORCL
🌐 Deploying to Hugging Face Spaces
Space ▸ Create ▸ Docker template
Push repo (Dockerfile, app.py, tools/, requirements.txt)
Add a secret → GEMINI_APIKEY
Space builds ; enjoy!
🗺 Project Structure
Copy code
├─ app.py
├─ requirements.txt
├─ Dockerfile
├─ tools/
│ ├─ csv_parser.py
│ ├─ plot_generator.py
│ ├─ forecaster.py
│ ├─ visuals.py
│ └─ db_connector.py
└─ README.md
🤝 Contributing
Issues & PRs are welcome!
Please run black -l 88 and ruff before submitting.
📜 License
MIT © 2025 BizIntel AI Ultra Team |