TuringArte

Knowledge (Knolo)

Ground AI Employees in your product docs, policies, and FAQs with Knolo packs.

Why Knolo?

TuringArte uses @knolo/core packs — local-first, deterministic knowledge artifacts built from your uploaded documents. Attached packs let the agent retrieve relevant chunks at chat time so answers stay on your content.

Workflow

  1. Open Dashboard → Knowledge.
  2. Upload documents (PDF, DOCX, Markdown, plain text, and more).
  3. Optionally set metadata tags and organize with LLM-assisted structuring (Knolo depth features).
  4. Create a pack and run build / rebuild.
  5. On the employee page, attach the pack under Knowledge so the runtime can query it.

What to upload

  • Product FAQs and troubleshooting guides
  • Pricing, plans, refunds, and policies
  • Hours, location, services (Receptionist)
  • ICP, case studies, value props (Sales)
  • Brand voice and content pillars (Content scaffold)

Presets suggest topic lists — see the agent catalog.

Best practices

  • Prefer clear section headings; avoid huge binary-only scans without text
  • Rebuild packs after significant doc changes
  • Keep secrets out of knowledge docs (API keys, passwords)
  • Test grounded answers in the playground with questions only your docs can answer

Embed signal

Public chat responses include usedKnowledge so you can verify retrieval is engaged in integrations.

Next: Tools & channels