Barrel
FR GitHub

Run Barrel in production with Enki Bridge

Fixed-scope pilots, managed hosting and support for private AI, RAG, agent memory and edge or local-first systems built on Barrel.

Contact Enki BridgeRead the docs

Offers

Built on the open source database. Start with one, grow into the rest.

Private RAG Pilot

Deploy Barrel DB for private semantic search and RAG over your own data. We help with ingestion, embeddings, hybrid search, metadata filtering, evaluation, and deployment.

  • Barrel Vector + documents under one id
  • local, Ollama, or OpenAI-compatible embeddings
  • hybrid vector/BM25 search
  • evaluation and tuning
Discuss a RAG pilot

Agent Memory Pilot

Use Barrel as a persistent memory layer for AI agents and MCP tools. Store documents, vectors, blobs, sessions, and capabilities behind one API.

  • REST and MCP access
  • spaces, sessions, handoffs
  • secure capability model
  • observability-ready deployment
Discuss agent memory

Edge Sync / Local-first Pilot

Build offline-first or edge deployments with sync, versioning, and local search. Keep data close to users and sync when connectivity allows.

  • document sync and conflict handling
  • local vector search
  • browser/edge deployment patterns
  • encrypted local storage
Discuss edge sync

Managed Barrel DB

Run Barrel DB with operational help from Enki Bridge. Available for selected design partners and production users.

  • deployment and upgrades
  • monitoring and backups
  • security review
  • private support channel
  • custom SLAs
Ask about managed Barrel

Open source core

Barrel is Apache 2.0 licensed. You can self-host it, embed it, fork it and run it without usage fees. Commercial support is for teams that want help with production architecture, deployment, operations or dedicated support.

Who this is for

  • European teams building private AI systems
  • Erlang/Elixir teams that want embedded vector/document storage
  • companies that cannot send data to a third-party vector SaaS
  • edge/local-first applications
  • teams exposing data to agents through REST or MCP

How it works

  1. Architecture call

    A short call to understand what you are building, your data, and your constraints. No cost, no obligation.

  2. Fixed-scope pilot or production review

    A design-partner engagement with a clear deliverable: a RAG pilot, an agent memory layer, an edge deployment, or a review of an existing production setup.

  3. Deployment, support, or managed service

    Ongoing commercial support, help with your own deployment, or a selected managed Barrel deployment run by Enki Bridge.

Up and down to moveEnter to open