LatticeDB
Embedded property-graph database with vector search and full-text indexing
Visit latticedb.org ↗What is LatticeDB?
LatticeDB is a single-file embedded database designed for applications that need to query connected data by relationships, semantics, and text simultaneously. It combines graph traversal, HNSW vector similarity search, and BM25 full-text search in one query layer using Cypher syntax, with no server or configuration required. The project describes itself as built for relationship-heavy workloads on a single machine.
Features include ACID transactions, durable named streams with graph changefeeds, configurable vector indexing parameters, fuzzy search with Levenshtein distance, and a portable single-file format. Available as a CLI tool and via Python, Node.js, and Java bindings, it targets use cases like Graph RAG, agent memory systems, and local knowledge applications.
Momentum
Daily activity over the last 30 days: upvotes, unique visits, new GitHub stars and Hacker News points.
How do we know it's vibe coded?
vibetrend discovered LatticeDB automatically. Evidence: GitHub jeffhajewski/latticedb; repo has AGENTS.md. View source.
Similar vibe-coded apps

Real-time global intelligence dashboard with live geopolitical signals


Privacy-first web analytics alternative to Google Analytics

Open container format for 4D gaussian splat scenes with spatial audio

Open-source, self-hosted logs and traces with full-text search on object storage

Rank the top 50,000 GitHub developers by lifetime contributions


A-share financial data API for AI agents, quant research, and developers