AI Agent ReadyConcurrentArrow Flight SQL

    A shared data platform your AI agents can query concurrently.

    DuckDB is great in-process, but agents working together need a shared, concurrent data platform. GizmoSQL is a lightweight, high-performance data platform that speaks Arrow Flight SQL, so many LLM agents query at once — with a Python ADBC driver that drops into LangChain, CrewAI, and your own frameworks.

    Concurrentmulti-agent access
    ADBCPython driver
    $0.17per TPC-H 1 TB run
    Install Core (free)

    Instant key. No sales call. No card. Backed by a 30-day money-back guarantee.

    100K+ Docker pullsIn production at Summation & RandoliFeatured on Product Hunt
    gizmosql_server
    $ gizmosql_server --flight :31337
    py> from adbc_driver_flightsql import connect
    py> db = connect('grpc://gizmo:31337')
    agent_a ── query ─┐
    agent_b ── query ─┼─▶ one shared platform
    agent_c ── query ─┘

    Concurrent access

    Multiple agents query the same data platform at the same time over Arrow Flight SQL.

    Drops into your framework

    Python ADBC driver for LangChain, CrewAI, and whatever you are building.

    Secure by default

    TLS and auth built in. Per-catalog permissions and multi-tenancy on Enterprise.

    Install Core, free

    Native binaries or Docker. Start the data platform in minutes.

    Point it at your data

    Parquet, Iceberg, Delta, or DuckLake on your object storage. No copies.

    Connect your agents

    One ADBC connection string, shared across every agent in the workflow.

    GizmoSQL for AI agents — FAQ

    Why not just use DuckDB in-process for my agents?

    In-process DuckDB binds the database to a single process, so concurrent agents fight over file locks or need fragile copies. GizmoSQL wraps DuckDB in a server that speaks Apache Arrow Flight SQL, giving every agent shared, concurrent access to one platform — with TLS and authentication built in.

    How do my agents connect to GizmoSQL?

    Through the Python ADBC Flight SQL driver — one connection string shared across every agent. It drops into LangChain, CrewAI, and custom frameworks, and JDBC and ODBC drivers cover everything else.

    Is GizmoSQL secure enough for production agent workloads?

    Yes. TLS and authentication are built into Core, and GizmoSQL Enterprise adds per-catalog permissions and multi-tenancy so each agent or tenant sees only the data it should.

    What does GizmoSQL cost?

    The core is Apache 2.0 open source and free — on the reproducible TPC-H 1 TB benchmark a full run costs about $0.17 in compute. Enterprise features are priced per vCPU, with a free 30-day trial and an instant key.

    Give every agent in the workflow one fast, shared data platform.

    Instant key. No sales call. No card. Backed by a 30-day money-back guarantee.