Cogwit runs Cognee’s pipelines in the cloud, but without the need to install databases, manage schedulers, or configure LLM keys. You can use Cogwit as the simplest way to start with Cognee, or connect it with a local setup for persistent, shared, always-on workloads.

Key capabilities

  • Managed infrastructure — Scaled Kuzu, LanceDB, and PostgreSQL clusters without running servers yourself.
  • Modal orchestration & pipeline automation — Add, cognify, and memify pipelines are preconfigured in Modal and ready to run, so you can use them immediately without setup.
  • Cogwit UI — A notebook-based UI for uploading files, monitoring runs, and searching memory.
  • Cogwit SDK — Python SDK interface that matches open source Cognee, secured with API keys.

Cognee OSS vs Cogwit

  • Cognee OSS — Run locally for experimentation, full control, or when you want to tinker with every part of the stack.
  • Cogwit — Run in the cloud for simplicity, persistence, and collaboration without operational overhead.
  • Combine both — Use Cognee locally for fast iteration, and Cogwit in the cloud for production pipelines and shared datasets.

When Cogwit fits

  • You want to try Cognee without setting up local infrastructure
  • Your pipelines need to stay online for continuous ingest and search
  • Your team requires secure, multi-tenant access with audit trails
  • Your workflow mixes interactive knowledge building with automated jobs

Get started

  • Sign up & prerequisites — Create your account, subscription, and API key to begin using Cogwit.
  • Cogwit UI — Upload a dataset and explore it through the hosted notebook interface.
  • Cogwit SDK — Automate ingestion and search using the Python SDK.

Learn more

  • Cogwit architecture — See how Cogwit components (orchestration, storage, UI, APIs) fit together.
  • Permissions & security — Learn how Cogwit handles authentication, multi-tenancy, and audit trails.
  • Local mode & sync — Connect to local Cognee instances and sync data between environments.