DiscovAI Search — Open‑Source AI Search Engine for Tools, Docs, and Custom Data

Source: Dev.to

Overview

DiscovAI Search is an open‑source, AI‑powered search engine designed to index, understand, and search AI tools and custom knowledge bases using modern vector search combined with LLM reasoning.

Key Features

• Semantic search with embeddings
• LLM‑powered answer generation
• Fast caching with Redis
• Structured storage using Supabase (PostgreSQL + pgvector)
• Modern frontend built with Next.js
• Fully open source

Suitable Use Cases

• Production‑ready AI search layer
• Educational reference project for AI and web developers

Architecture Diagram

User Query
   ↓
Next.js API Route
   ↓
Embedding (OpenAI)
   ↓
Vector Search (Supabase / pgvector)
   ↓
Redis Cache (optional)
   ↓
LLM‑generated response
   ↓
UI

Design Benefits

• Scalable
• Modular
• Easy to extend with new data sources

Technology Stack

Prerequisites

• OpenAI API key
• Supabase account
• Redis instance (local or cloud)

Installation

git clone https://github.com/DiscovAI/DiscovAI-search.git
cd DiscovAI-search

# Using npm
npm install

# Or using yarn
yarn install

Environment Variables

Create a .env.local file with the following keys:

OPENAI_API_KEY=your_openai_key

NEXT_PUBLIC_SUPABASE_URL=your_supabase_url
NEXT_PUBLIC_SUPABASE_ANON_KEY=your_supabase_anon_key

SUPABASE_SERVICE_ROLE_KEY=your_service_role_key

REDIS_URL=redis://localhost:6379

Supabase Setup

1. Enable the pgvector extension.
2. Create tables for documents and embeddings, storing vectors in vector columns.

Run the Development Server

npm run dev   # or `yarn dev`

Open http://localhost:3000 in your browser to see the DiscovAI Search interface.

Indexing Data

DiscovAI Search can index:

• AI tools
• Documentation
• Articles
• Internal knowledge bases

Typical flow:

1. Add documents to Supabase.
2. Generate embeddings via OpenAI (or an alternative model).
3. Store vectors in the pgvector column.
4. Query through the UI.

Extending the Project

• Add your own datasets.
• Swap OpenAI embeddings for open‑source models.
• Connect multiple vector indexes.
• Add authentication.
• Deploy to Vercel (frontend) with Supabase (database) and Upstash Redis (cache).

Why DiscovAI Search Matters

• Demonstrates a real‑world AI search architecture.
• Correctly combines LLMs with vector databases.
• Easy to fork, customize, and deploy.
• Serves as both a product and a reference implementation.

Potential Applications

• AI‑powered search engines
• Internal knowledge assistants
• Tool discovery platforms

Further Exploration

If you’re interested in:

• Semantic search
• Vector databases
• LLM‑powered user experiences

…DiscovAI Search is a solid starting point.

Repository

https://github.com/DiscovAI/DiscovAI-search/