lance

Browse Lance tables from your local machine in a simple web UI. No database to set up. Mount a folder and go.

Integration between Lance and Ray for distributed data processing

Comparing LanceDB and Elasticsearch for full-text search and vector search performance

Incremental processing and maintaining data freshness with CocoIndex and LanceDB

Cocoindex-based code indexing service with HNSW, automatic disk compaction, MCP Server, periodic git pull, metrics push to MySQL, and space-efficient branch searching

Inspect, query, and export Lance datasets from the command line

Business data as geometry that AI agents navigate: explainable anomaly detection, drift and chain investigation with no model training. Snapshot sessions, copy-free branches, Lance storage, MCP-ready.

Using the Lance extension in DuckDB for querying, joining and materializing on tables in either system

✨ Un jeu du dé totalement fait en Python par moi, et en français.

A minimal FastAPI API that accepts an image via POST, stores it locally with a SHA256 filename, vectorizes it using OpenAI's CLIP model, and appends the vector to a local Lance dataset. Designed as a small MVP for local or Fly.io deployment.

A distributed data-processing engine for large-model (LLM) data — SQL, curation, vector/FTS retrieval, multimodal — on stock DuckDB with Rust orchestration.

Minimal FastAPI project to test storing and searching image vector data using the Lance format on Cloudflare R2.

Pack dataset media into plain STORED zip shards; read them fast anywhere.

Chat-style multimodal assistant for ByteDance Lance (T2I/T2V/edit/understand) with optional OpenAI-compatible VLM orchestrator. FastAPI + Preact streaming UI, auto-downloading weights.

Lightweight ML experimentation toolkit: versioned datasets, reproducible runs, experiment tracking, cloud GPUs. No servers.

Convert Beneficial Ownership Data Standard (BODS) v0.4 data to the Lance columnar format

An operator-focused control plane for governed vector data: versioned documents in Lance, catalog & lineage in DuckDB, serving in Qdrant, and quality-gated embedding migrations with zero-downtime cutover.

Business data as geometry that AI agents navigate: an MCP server for anomaly, drift, root-cause and chain investigation. Stdio or local HTTP, snapshot sessions, branches, compact JSON for agent context.