Open source and managed cloud
Inquisitive Fox, Copernicus, Zebra Pudding, and Sylvester's Hat each ship as open source. Extend them and pull them into your own projects. There is a managed cloud when you want hosted access.
What we build
Ideas we own and operate
Other solutions were already out there. That is not why we built these. We needed them to be different. Sometimes that means privacy. Sometimes it means how AI is managed. Sometimes it is a development priority that off-the-shelf tools will not bend to.
Adapters, not lock-in
Those open source systems use adapters so you can swap backends and third-party services without rewriting the product. Integration and privacy choices stay under your control.
Architect-directed build
We use current foundation models to produce the code. Software architects direct that work so security, compliance, breadth, and reliability stay intact.
Your company brain
A private company brain that extracts intelligence from the documents, chats, and files your team already has. Connect Slack, Google Drive, Gmail, Microsoft 365, and more. Ask a question, get a cited answer, and let agents take the next step.
Inquisitive Fox
Coming soonLong-term memory for AI
Long-term memory for AI applications and agents. Pluggable and provider-agnostic, so you can extend the open source system, drop it into your own projects, or run the managed cloud. Adapters let you swap backends and third-party systems without rewriting the product.
Copernicus
Coming soonSmart prompt retrieval
Retrieve the right prompt for the job across models and other attributes. Version prompts, run evaluations, and assemble what your applications and agents need at runtime. Open source with adapters for backends and third-party systems, plus a managed cloud when you want hosted access.
Zebra Pudding
Coming soonLearning and development with courses and coaching
A learning and development system that puts virtual courses and coaching in one place. We did not bolt AI onto a generic LMS. We needed different choices around privacy, how AI shows up in learning workflows, and product priorities that off-the-shelf stacks would not bend to. Available as open source and as managed cloud.
Sylvester's Hat
Coming soonSmart LLM routing
Route each request to the right model based on attributes like latency, cost, context size, intelligence, and mode (text, images, video, audio). Query categorization helps pick the path before a model is chosen. Open source with adapters so you can swap backends and providers, plus managed cloud for hosted access.