Supavec provides an open-source RAG-as-a-Service platform, enabling users to build Retrieval-Augmented Generation applications with diverse data sources. It targets developers and researchers seeking a scalable, self-hostable alternative to commercial solutions like Carbon.ai, offering a rapid deployment option for vector search and chat APIs.

How It Works

Supavec employs a multi-tenant architecture with row-level security for data isolation. It optimizes RAG performance through configurable chunking and overlap strategies, hybrid filtering (combining file ID and cosine similarity), and asynchronous processing for embeddings. This approach aims to reduce costs, improve recall, and maintain low latency for vector search operations.

Quick Start & Requirements

• Install dependencies: bun i
• Run development server: bun dev
• Prerequisites: Bun runtime.
• Further details: docs.supavec.com

Highlighted Details

• Offers usage-based billing with tiered plans.
• Achieves P95 latency of 210ms for vector search.
• Integrates PostHog for real-time analytics and request-level tracing.
• Features a chat UI with live embedding previews for visual debugging.

Maintenance & Community

The project is actively developed, with a presence on Product Hunt and GitHub. Links to related repositories and API documentation are provided.

Licensing & Compatibility

The repository does not explicitly state a license in the provided README. Users should verify licensing for commercial or closed-source integration.

Limitations & Caveats

The README does not detail specific limitations, unsupported platforms, or known bugs. The project appears to be relatively new, indicated by its Product Hunt launch.