# Agami

> The trust layer for your data. Auto-generated semantic models, vetted by your data team, and answered from every tool your team already uses (Claude, Slack, Gemini, Notion, custom agents). Built by data engineers and analysts who shipped this internally at FANG-scale companies.

## What Agami does

Agami sits between your AI tools and your data. It auto-generates a semantic model from your existing schema, has your data team vet every definition, applies your row/column security and SSO rules, and answers business questions consistently from every assistant your team uses.

Three properties that make Agami different from a plain LLM-on-warehouse setup:

1. **Auto-authored** — the semantic model is generated from your schema and kept in sync as the database evolves.
2. **Human-vetted** — your data team reviews and approves every definition before it ships. The semantic model, not the LLM, picks the formula, grain, and joins.
3. **Served everywhere** — the same governed answer in Claude, ChatGPT, Slack, Notion, Gemini, Cursor, and custom agents over MCP.

## Pages

- [Home](https://www.agami.ai/): product overview and positioning
- [How it works](https://www.agami.ai/how-it-works): runtime path of a single question through the trust layer
- [About](https://www.agami.ai/about): founders (Sandeep Kachru, Co-founder & CEO; Ashwin Ramachandran, Co-founder & CPTO) and why Agami exists
- [Open source](https://www.agami.ai/oss): LiteBi, the open-source distribution of Agami, the trust layer between AI and your data (Apache-licensed, runs locally)
- [FAQ](https://www.agami.ai/faq): common questions about Agami, the trust layer, and integrations
- [Free trial](https://www.agami.ai/free-trial): start a trial with your own data
- [Talk to us](https://www.agami.ai/talk-to-us): book a 30-minute conversation with the founders

## Deployment options

- **Local (open source)** — Run Agami's LiteBI locally on your machine. Apache 2.0 licensed. No Agami server, no API key. https://github.com/AgamiAI/LiteBi
- **Teams (hosted)** — Agami hosts your team's data definitions. Access through the Agami MCP Server. Integrations for every tool and assistant. Bring your own LLM.
- **Enterprise** — Dedicated tenant. Detailed audit trails. Priority support. Setup engagement included. Built for enterprise-level compliance.

## Data sources

Connects in place to Snowflake, Databricks, Postgres, MySQL, Redshift, BigQuery, and other warehouses your team already uses. No data movement or copying.

## Contact

- Email: founders@agami.ai
- LinkedIn (company): https://www.linkedin.com/company/agami-ai/