Show HN: Omni – Open-source workplace search and chat, built on Postgres

Source: Hacker News

Overview

Over the past few months I’ve been building Omni, an open‑source workplace search and chat platform that connects to apps like Google Drive/Gmail, Slack, Confluence, Jira, HubSpot, and more. It’s intended as a self‑hosted alternative to Glean for small‑ to mid‑size teams, running entirely on PostgreSQL (ParadeDB) and pgvector—no Elasticsearch or dedicated vector databases.

Getting Started

Deploy Omni on your own infrastructure with a single command:

docker compose up

Then configure your apps and LLM provider (OpenAI, Anthropic, Gemini, etc.) through the UI.

Features

• Data sync from connected apps with a BM25 index (ParadeDB) and an HNSW vector index (pgvector).
• Hybrid search that combines BM25 and vector results.
• Chat UI where the LLM can invoke search tools (beyond basic RAG).
• Traditional search UI for keyword queries.
• Bring‑your‑own LLM support (OpenAI, Anthropic, Gemini, etc.).
• Connectors for Google Workspace, Slack, Confluence, Jira, HubSpot, and more.
• Connector SDK to build custom integrations.

Feedback Requested

Omni is currently in beta. I’d appreciate any input on:

• Experiences with self‑hosting workplace search or AI tools.
• Concerns about a PostgreSQL‑only architecture at larger scales.

Feel free to ask questions!

Code: (Apache 2.0 licensed)
Comments: