What Is Reticuli

Understand what Reticuli is, what it helps you build, and which page to read next.

Reticuli

Reticuli is a product for building AI assistants from your own knowledge, examples, and deployment needs.

Use it when you want an assistant that can answer from your documents, follow your preferred style, test replies before launch, and move toward a live endpoint when the result is ready.

Build, test, deploy, review
Reticuli production workflow from trusted material to deployment and usage review.

Reticuli combines source grounding, training examples, playground testing, deployment keys, and workspace usage review in one project workflow.

What you can build

  • A support assistant that answers from help docs and previous tickets.
  • An internal assistant that understands policies, SOPs, handbooks, and playbooks.
  • A product assistant that cites the right documentation before answering.
  • A drafting assistant that follows examples from your strongest team members.
  • A private assistant that combines retrieval, fine-tuning, testing, and deployment in one project.

The basic idea

Reticuli has three practical building blocks:

Building blockPlain meaningUse it when
Knowledge baseDocuments the assistant can search before answeringAnswers must use current facts, policies, or product details
Fine-tuningExamples that teach tone, structure, and judgmentAnswers need to sound and behave like your team
DeploymentThe live assistant endpoint and keysUsers or systems need to use the assistant outside setup screens

Every project can use source data, training examples, playground tests, and deployment. You do not need to choose a narrow project type up front.

Real workflow

  1. Sign in with passkey, GitHub, Google, or email code.
  2. Create a project with one assistant job.
  3. Add trusted data on the canvas.
  4. Review source files and generated examples before training.
  5. Test with the same prompt set in Playground.
  6. Deploy through API key when answers are stable.
  7. Watch usage and update sources or examples as real questions arrive.

Reading order

If you are new, read these pages in order:

  1. Get Started
  2. Product map
  3. Create a project
  4. Knowledge base
  5. Playground and testing
  6. Deploy an assistant

If you already understand AI products, go straight to Advanced usage.