Show HN: HelixDB – A graph database built on object storage
Graph databases just got cheaper. HelixDB runs on S3, not RAM—enabling TB-scale AI memory for 1/10th the cost.
Why it matters
HelixDB solves a critical infrastructure problem for AI agents: combining graph, vector, and full-text search in one system while using object storage instead of expensive distributed replicas. For founders building agentic systems, this eliminates the tax of stitching together 3+ databases.
The key facts
6 to knowHelixDB built on object storage (S3) rather than distributed in-memory replicas
Supports TB-scale graph data with p99 ~100ms writes, ~50ms reads from cold storage
Hybrid graph/vector/FTS architecture—native joins across all three modalities
Targets AI agent memory systems and company knowledge graphs
Open-source with cloud GA launching within weeks
Solves graph sharding problem (edges span partitions) by caching hot subsets on nodes
Go to the source
Hacker Newsgithub.com
Publisher excerpt: Hey HN, it’s been just over a year since we launched HelixDB (https://news.ycombinator.com/item?id=43975423), a project a friend and I started in college. It’s an OLTP graph database built on object-storage, with native vector search and full-text search (FTS). Why graph, vector and FTS? Graph…