From raw documents to production-ready AI agents.
Nata handles chunking, embeddings, RAG pipelines and streaming responses, so you can focus on building the experience.
Train with your data
Upload PDFs, docs or text. We chunk, embed, and set up RAG automatically.
Multiple Nats per workspace
Spin up different agents for support, docs, analytics or internal tools — each with its own brain.
One API, many surfaces
Use your Nat in a web widget, inside your app, or via REST API. Same brain everywhere.
A single endpoint for all your Nats.
Call your Nat from any backend or frontend. We handle RAG, prompts, and streaming under the hood.
curl -X POST https://api.nata.ai/public/chat \
-H "X-API-Key: $NATA_PUBLIC_KEY" \
-H "Content-Type: application/json" \
-d '{
"workspace_id": 1,
"bot_id": 2,
"session_id": "demo-user",
"message": "Summarize my onboarding docs"
}'Frequently asked questions
Quick answers about Nats, data, and getting started.
What is a Nat?+
A Nat is an AI agent trained on your documents with memory, tools, and a purpose. You can run multiple Nats in the same workspace.
Do I need to build my own RAG pipeline?+
No. Nata handles chunking, embeddings, and retrieval automatically so you can focus on the experience.
Can I use my own data securely?+
Yes. You control what gets uploaded, and each bot only accesses the data you provide.
How fast can I launch?+
Most teams can create a workspace, upload docs, and deploy a bot in minutes.
Ready to spin up your first Nat?
Start with a single workspace, connect your documents, and launch an AI agent in minutes — without managing any infra.