Equity Engine
A conversational AI specialized in workplace fairness, with private per-client knowledge bases.

How it works.
A Fair360 consultant creates a private research 'folder' for a client company, industry, or from scratch, each backed by its own OpenAI vector store and GCS-stored company data. Inside a folder they hold multi-turn chats, streamed over SSE, with an AI persona specialized in workplace-fairness/DEI analytics that refuses off-topic queries. The assistant can run domain-whitelisted web search, search the folder's uploaded files, and execute pandas code in a Docker sandbox against the company's data. Consultants can share a chat read-only with colleagues, and the assistant remembers user preferences and past insights across sessions.
- 01Creates per-company private folders, each backed by its own OpenAI vector store and GCS data
- 02Streams multi-turn chat via SSE using the OpenAI Responses API
- 03Runs domain-whitelisted web search, vector file search, and sandboxed pandas code execution as chat tools
- 04Shares chats read-only with colleagues; retains cross-session memory of user preferences
How to use Equity Engine.

Where it stands.
Next phases
Built with.
Next.js 16 + React 19FastAPI + SQLAlchemy 2PostgreSQL 15 + Redis 7Want Equity Engine working for you?
A conversational AI specialized in workplace fairness, with private per-client knowledge bases.
Get in touch