Vekku - AI powered content storage with smart tag suggestions

kuchuk-borom-debbarma — Sun, 01 Feb 2026 08:27:18 +0000

https://vekku-reactjs.vercel.app/

Vekku

Transforming raw data into a structured semantic web.

Vekku is an intelligent digital knowledge base designed to help you organize your thoughts, projects, and ideas with the power of AI. Unlike traditional note-taking apps that rely on manual organization, Vekku uses semantic analysis to automatically suggest tags, discover connections between concepts, and help you build a graph of knowledge effortlessly.

🚀 Why Vekku?

🧠 Semantic Intelligence: Automatically analyzes your content to suggest relevant tags and categorize information based on meaning, not just keywords.
⚡ High Performance: Built on the Bun runtime and React 19, ensuring instant interactions and lightning-fast data processing.
💎 Glassmorphism UI: A beautiful, modern interface featuring liquid-glass aesthetics that provides a focused and pleasant user experience.
🌐 Platform Agnostic Backend: Architected to run anywhere — from a local high-performance server to edge networks like Cloudflare Workers.

🛠️ Tech Stack

Vekku is a full-stack application composed of two main parts:

Frontend (`/web`)

Framework: React 19 (via Vite)
Language: TypeScript
Styling: Tailwind CSS v4, shadcn/ui (Radix UI)
State: React Context & Hooks
Design: Modern glassmorphism with responsive layouts

Backend (`/backend`)

Runtime: Bun (Optimized for speed)
Framework: Hono (Platform agnostic adapter pattern)
Database: PostgreSQL (via Neon Serverless)
ORM: Drizzle ORM
AI: Cloudflare AI (Text embedding models)
Architecture: Event-driven, functional modules

📂 Project Structure

Vekku-Meta/
├── backend/    # The API service (Bun/Hono/Drizzle)
├── web/        # The Frontend application (React/Vite)
├── docs/       # Documentation and architectural decisions
└── README.md   # This file

📚 Documentation

For deeper dives into specific parts of the system:

Backend Guide: Detailed API flows, auth strategy, and pagination logic.
Frontend Guide: UI architecture, component structure, and state management.
Docs Folder: Architectural decisions and future roadmap.

💻 Local Development Setup

Vekku is already deployed and ready to use. However, if you wish to run a local instance or contribute to development, follow the steps below.

Prerequisites

Node.js & npm (for Frontend)
Bun (for Backend)
PostgreSQL Database (e.g., Neon)
Cloudflare Workers AI (for Embeddings)
Upstash Redis (for Caching/Rate Limiting)

1. Backend Setup

Navigate to the backend directory:

cd backend
bun install

Environment Variables:
Copy .env.example to .env and populate the following secrets:

Variable	Description
`DATABASE_URL`	PostgreSQL connection string (Neon Tech).
`JWT_SECRET`	Secret key for signing authentication tokens.
`CLOUDFLARE_WORKER_AI_API_KEY`	API Key for Cloudflare Workers AI.
`CLOUDFLARE_WORKER_ACCOUNT_ID`	Cloudflare Account ID.
`CLOUDFLARE_AI_MODEL`	Embedding model (e.g., `@cf/baai/bge-small-en-v1.5`).
`EMBEDDING_THRESHOLD`	Similarity threshold for suggestions (e.g., `0.4`).
`EMBEDDING_MATCH_COUNT`	Number of suggestions to return (e.g., `10`).
`NOTIFICATION_API_CLIENT_ID`	Client ID for notification service (optional).
`NOTIFICATION_API_CLIENT_SECRET`	Secret for notification service (optional).
`UPSTASH_REDIS_REST_URL`	URL for Upstash Redis instance.
`UPSTASH_REDIS_REST_TOKEN`	Auth token for Upstash Redis.
`FRONTEND_URL`	URL of the frontend (e.g., `http://localhost:5173`).

Run the development server:

bun run dev

2. Frontend Setup

Navigate to the web directory:

cd web
npm install

Environment Variables:
Create a .env file with the following configurations:

Variable	Description
`VITE_API_URL`	Full URL to the backend API (e.g., `http://localhost:3000/api`).

Start the frontend development server:

npm run dev

The application should now be running at http://localhost:5173.

Created by **Kuchuk Borom Debbarma**

DEV Community: kuchuk-borom-debbarma