apocas/restai
RESTai is an AIaaS (AI as a Service) open-source platform. Built on top of LlamaIndex & Langchain. Supports any public LLM supported by LlamaIndex and any local LLM suported by Ollama/vLLM/etc. Precise embeddings usage and tuning. Built-in image generation (Dall-E, SD, Flux) and dynamic loading generators.
What's novel
RESTai is an AIaaS (AI as a Service) open-source platform. Built on top of LlamaIndex & Langchain. Supports any public LLM supported by LlamaIndex and any local LLM suported by Ollama/vLLM/etc. Precise embeddings usage and tuning. Built-in image generation (Dall-E, SD, Flux) and dynamic loading generators.
Code Analysis
4 files read · 2 roundsA full-stack React/Python application that orchestrates multiple LLMs and embeddings to route user queries to specific AI models based on semantic analysis.
Strengths
Strong architectural separation between frontend (React/MUI) and backend (FastAPI/SQLAlchemy). The database schema is well-designed for managing complex relationships between users, teams, projects, and AI models. The initialization logic provides a robust way to seed the system with demo data and default configurations.
Weaknesses
Lack of visible unit tests or integration tests in the provided snippets. Error handling appears basic (standard FastAPI exceptions) without custom error pages or detailed logging strategies shown. The frontend relies heavily on Context API which can lead to prop-drilling issues if not managed carefully at scale.
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
99
Contributors
7
Files
350
Active weeks
17
Repository
Language
JavaScript
Stars
482
Forks
95
License
Apache-2.0