AI & ML interests

Synthetic data, Conversational AI, Persona simulation, Computational social science, LLM agents

Recent Activity

echosim  updated a Space 21 days ago
echo-sim/README
echosim  published a Space 21 days ago
echo-sim/README
echosim  updated a Space 21 days ago
echo-sim/echosim-first-contact-demo
View all activity

Organization Card

EchoSim

Research artifacts for simulation-first compatibility matching.

EchoSim studies a simulation-first approach to dating: instead of pushing two strangers through weeks of low-signal messaging, each person is represented by an AI persona (an "Echo") that pre-simulates first-contact conversations to estimate compatibility before any human interaction begins.

This organization publishes the open, non-sensitive artifacts of that work — synthetic persona dialogues and conversational-AI resources. We deliberately do not release production matching logic, scoring models, or any user-derived data.


What we publish

  • 📊 Datasets — fully synthetic persona-to-persona dialogue corpora (no real users, no PII).
  • 🧪 Spaces — small interactive demonstrations of the simulation-first concept.

Research focus

Conversational AI · persona simulation · synthetic dialogue generation · computational social science · human–AI interaction.

Scientific grounding

Persona modeling draws on widely used descriptive frameworks — adult attachment theory, MBTI communication styles, and the Big Five (OCEAN) traits. These are used as descriptive tags for research, not as clinical or predictive claims.

Privacy & ethics

All published data is 100% synthetic, generated from sampled persona attributes. It contains no real conversations and no personally identifiable information. These artifacts are intended for research, prototyping, and education — not for real-world relationship, psychological, or hiring decisions.

Links

Content released under CC-BY-NC-4.0 unless noted otherwise.

Edit this README.md markdown file to author your organization card.

models 0

None public yet