apocas/restai

RESTai is an AIaaS (AI as a Service) open-source platform. Supports many public and local LLM suported by Ollama/vLLM/etc. Precise embeddings usage, tuning, analytics etc. Built-in image/audio generation with dynamic loading generators. Live chat deployment. Built-in block based graphical language. Prompt versioning and much more...

RESTai – AI‑as‑a‑Service platform

What it is – A self‑hosted web service that lets you create, manage and expose AI projects (LLM chat, RAG, agents, image generation, etc.) through a simple REST API and a built‑in React dashboard. It bundles everything needed to run LLM‑backed applications without writing code: data ingestion, vector stores, prompt versioning, analytics, multi‑tenant teams, and optional Kubernetes/Docker deployment.

Core capabilities

  • Multi‑project management – each project can pick its own LLM (OpenAI, Anthropic, Ollama, Gemini, Azure, LiteLLM, vLLM, …) and configuration.
  • RAG & Knowledge Graph – upload docs, sync from URLs/S3/Confluence/SharePoint/Google Drive, store in vector DB, optionally run NER to build an entity graph for richer retrieval.
  • Agents & MCP – zero‑shot ReAct agents with built‑in tools (browser automation via Playwright, terminal, classifier, etc.) and support for the Model Context Protocol to expose custom tools.
  • Visual Block builder – Blockly‑style flow editor for composing pipelines without code; can call other RESTai projects.
  • Analytics & cost tracking – dashboards show token usage, latency, per‑project spend, and guard‑rail statistics.
  • Security & governance – RBAC, OAuth/LDAP, TOTP 2FA, encrypted secrets vault, per‑project rate limits, input/output guards, audit log.
  • Direct OpenAI‑compatible endpoints – /v1/chat/completions, /v1/images/generations, /v1/audio/transcriptions work out‑of‑the‑box.
  • Deployability – install via PyPI (restai-core), Docker image (apocas/restai), or Kubernetes; auto‑detects GPUs and installs the right dependencies.

Typical use cases

  • Internal knowledge‑base chatbot for a company (RAG with document sync).
  • Customer‑support widget embedded on a website.
  • Automated agents that log into vendor portals, scrape data, and generate reports.
  • Multi‑team SaaS offering where each client gets its own branded AI console.
  • Rapid prototyping of AI pipelines using the visual Block IDE.

Getting started

pip install restai-core          # installs backend + React UI
restai init                      # creates DB & admin user
restai migrate                   # apply migrations
restai serve -p 8080 -w 4        # launch http://localhost:8080/admin

Or run the official Docker image:

docker run -p 9000:9000 apocas/restai:latest

Both approaches give you a full UI (admin/admin) and the REST endpoints.

License – Apache 2.0 (per the badge in the README).

All information above is taken directly from the repository’s README; no additional features are inferred.