Production-Grade Database Features

Columnist-DB combines columnar storage, vector search, and AI memory in a client-side database. Built for performance, type safety, and developer experience.

Key Highlights

Query 100k embeddings in <50ms
Local-first with IndexedDB
Full TypeScript support
◇ Feature Overview

Columnar Storage

Optimized for analytical queries and vector operations. Data is stored column-by-column for better compression and query performance.

Vector Embeddings

Store and query high-dimensional vectors with cosine similarity search. Perfect for AI applications, semantic search, and recommendation systems.

IndexedDB Persistence

All data persists locally in the browser with full offline support. ACID-compliant transactions ensure data integrity.

TypeScript First

Fully typed API with Zod schema validation. Type-safe migrations and automatic type inference from your schema.

Core Features

Everything you need for modern client-side data management

Columnar Storage

Data is stored column-by-column for optimal compression and analytical query performance. Perfect for aggregations and large datasets.

•Efficient compression algorithms
•Fast analytical queries
•Optimized for vector operations

Vector Search

Built-in vector embeddings support with cosine similarity search. Query 100k+ embeddings in milliseconds.

•Cosine similarity search
•Hybrid text + vector queries
•Semantic search capabilities

IndexedDB Persistence

All data stays local in your browser. No server required, complete offline support, and full privacy.

•ACID-compliant transactions
•Offline-first architecture
•Cross-session persistence

TypeScript First

Fully typed API with automatic type inference. Zod schema validation ensures type safety at runtime.

•Auto-generated types from schema
•Runtime validation with Zod
•Type-safe migrations

High Performance

Optimized for speed with columnar compression and efficient indexing. Handle millions of records with ease.

•100k embeddings in <50ms
•Smart indexing strategies
•Optimized vector operations

Schema Migrations

Version your database schema and migrate data safely. Full control over schema evolution.

•Version-controlled schemas
•Safe data migrations
•Rollback support

Advanced Capabilities

Full-Text Search

Built-in full-text search with relevance scoring. Search across multiple fields with fuzzy matching and field boosting.

db.fullTextSearch('memories', 'query', { fields: ['content', 'tags'], limit: 20 })

AI Memory Layer

Store conversations, documents, and knowledge with automatic embedding generation. Perfect for AI assistants and chatbots.

✓Cross-session memory persistence
✓Semantic search over conversations
✓16 MCP tools for integration

React Integration

Seamless React hooks for real-time data synchronization. Your components update automatically when data changes.

import { useQuery } from 'columnist-db-hook' function MyComponent() { const { data } = useQuery('memories') // Auto-updates on changes }

Query Builder

Powerful query API with filtering, sorting, pagination, and aggregations. Build complex queries with ease.

✓Complex filtering conditions
✓Multi-field sorting
✓Cursor-based pagination

Open Source

MIT licensed and fully open source. Contribute, fork, and use in any project without restrictions.

Framework Agnostic

Use with React, Vue, Svelte, or vanilla JavaScript. Works anywhere JavaScript runs.

Active Development

Regular updates with new features, performance improvements, and bug fixes. Community-driven development.