GitHub - bagofwords1/bagofwords: Chat with your data - with memory, rules, and observability built in. Deploy in 2 minutes

Features:

• Memory: Stores context, preferences, and usage patterns down to table/column level. AI learns over time and surfaces semantic-layer updates, review-gated.

• Rules: Define instructions, terms, and guardrails with versioning and approval workflows. Git sync with auto-indexing of dbt, markdown, code, and more.

• Observability: Full visibility into agent traces, plans, guardrails, LLM judges, evals, and feedback — debug and improve the loop.

• Dashboards & sharing: Build and share dashboards instantly. Share conversations, saved queries, and analysis with your team.

• Any LLM ↔ Any warehouse: OpenAI, Anthropic, Gemini, Ollama ↔ Snowflake, BigQuery, Postgres, Redshift. Swap or mix without breaking workflows.

• MCP: Use from Cursor, Claude Desktop, and other AI clients while reliably tracking requests and data operations.

Deploy anywhere

Quick Start 🚀

# runs with SQLite (default)
docker run -p 3000:3000 bagofwords/bagofwords

Or, run with a ready PostgreSQL instance

docker run -p 3000:3000 \
  -e BOW_DATABASE_URL=postgresql://user:password@localhost:5432/dbname \
  bagofwords/bagofwords

Other deployment options

docker compose and Kubernete deployments are also available (and recommended for servers). See the docs.

Product Overview

Chat with any data

Create reports, deep analysis or quick visuals with an AI interface powered by an agentic-loop with tools, reasoning and reflection built in.

Instructions Registry

Manage AI rules and instructions with review process, full versioning and tracking. Sync with git to auto-index dbt, Tableau, AGENTS.md and more.

Save data and queries to the Catalog

Leverage the catalog to store, share, and explore reusable queries and datasets. This feature also improves discoverability and searchability for AI, contributing to smarter AI decisions.

Monitor AI and data operations

Full observability into queries, feedback, and context — powering self-learning and high quality AI results

Architecture

Bag of words acts as a context-aware analytics layer that connects to any database or service, works with any LLM, and enriches queries with docs, BI models, or code.

The architecture is fully flexible: plug in any data source, any model, and any interface — giving your team maximum freedom of choice, without sacrificing governance or reliability.

Integrations

Supported LLM Integrations

Bag of words supports a wide range of LLM providers out of the box. You can bring your own API key for any of the following:

Tip: You can configure multiple providers and models, set defaults, and more.

Data Sources

Supported Data Sources

Below is a list of all data sources supported by Bag of words, as defined in the data source registry. Each entry is marked as either a Database/Warehouse or a Service.

Note: Some data sources (like NetSuite) may be marked as inactive or beta in the registry. "Service" refers to APIs or SaaS platforms, while "Database/Warehouse" refers to systems that store and query structured data.

Enterprise

For teams requiring advanced security, compliance, and governance:

• SSO: Google Workspace and OIDC-compatible identity providers
• Self-hosted: Deploy on your own infrastructure with full data control
• RBAC: Role-based access control for fine-grained permissions

🔒 Security & Privacy

We take data security and privacy seriously.

Telemetry

By default, Bag of words captures basic usage stats of self-hosted instances to a centralized server. The data helps us improve the product.

You can disable by setting in bow-config.yaml

You can also disable the Intercom chat for support