Launched this week
HeliosDB-Nano
Nano database with Vectors, Data Branching & Time‑Travel
1 follower
Nano database with Vectors, Data Branching & Time‑Travel
1 follower
HeliosDB-Nano is a single 32MB embedded SQL database with native PostgreSQL & MySQL wire‑protocol compatibility, built‑in vector search (HNSW + PQ), full‑text search, git‑like branching, and time‑travel queries. It runs anywhere — local‑first apps, edge devices, AI agents, or serverless. No migrations, no drivers to rewrite. Use psql, mysql, REST, or any ORM — everything just works.
Hey Product Hunt! We built HeliosDB-Nano because modern apps - especially AI‑powered and local‑first ones - need a database that’s fast, portable, and powerful without requiring cloud dependencies. HeliosDB-Nano gives you Postgres/MySQL compatibility, vector search, branching, time‑travel, and a built‑in BaaS layer in a tiny 32MB binary. Would love your feedback, ideas, and use cases!
Developers today juggle too many moving parts: a SQL database, a vector store, an auth service, a REST layer, and sometimes a proxy. We wanted a single embedded engine that could do all of this - locally, offline, or at the edge - without sacrificing SQL compatibility or performance.
Local‑first apps, AI agents, and edge workloads need a real database - not a toy key‑value store - but existing SQL engines are either too heavy, too limited, or lack vector search. We wanted a production‑grade SQL engine that works anywhere, speaks Postgres/MySQL, and includes modern features like vectors, FTS, branching, and time‑travel.
HeliosDB-Nano started as a lightweight embedded engine, but developers kept asking for Postgres/MySQL compatibility, vector search, and a built‑in BaaS layer. Instead of bolting features on, we redesigned the engine around RocksDB, Arrow, HNSW, and a unified wire‑protocol layer. The result is a tiny binary that behaves like a full database server.