DEV Community

Semurg profile picture

Semurg

Memory for AI agents: knowledge graph, vectors and graph RAG in one self-hosted engine, over MCP. First node free.

Joined Joined on  Personal website https://one.semurg.io
We kill -9 our database 21 times in 140 seconds, then read back every write. Here is what comes back

We kill -9 our database 21 times in 140 seconds, then read back every write. Here is what comes back

Comments
2 min read

Want to connect with Semurg?

Create an account to connect with Semurg. You can also sign in below to proceed if you already have an account.

Already have an account? Sign in
Give your LangChain, LlamaIndex, CrewAI, Google ADK, Hermes or OpenClaw agent a memory, one connection each

Give your LangChain, LlamaIndex, CrewAI, Google ADK, Hermes or OpenClaw agent a memory, one connection each

Comments
3 min read
We benchmark Semurg against the databases you already run: 11 domains, every answer checked

We benchmark Semurg against the databases you already run: 11 domains, every answer checked

Comments
3 min read
Loading a 117.2-million-relationship graph: Semurg 6.0 to 70.3 times faster than FalkorDB, ArcadeDB and SurrealDB

Loading a 117.2-million-relationship graph: Semurg 6.0 to 70.3 times faster than FalkorDB, ArcadeDB and SurrealDB

Comments
3 min read
A 1.8-billion-relationship graph in 12 GB of memory: Semurg answers in 1.41 s, Neo4j and FalkorDB did not finish

A 1.8-billion-relationship graph in 12 GB of memory: Semurg answers in 1.41 s, Neo4j and FalkorDB did not finish

Comments 3
3 min read
131,072 hops deep: Semurg walks a social graph up to 242 times faster than the incumbents

131,072 hops deep: Semurg walks a social graph up to 242 times faster than the incumbents

Comments
4 min read
Graph on one server: Semurg answers a 4-hop query 25 to 490 times faster than five graph databases

Graph on one server: Semurg answers a 4-hop query 25 to 490 times faster than five graph databases

Comments
4 min read
Semurg: local persistent memory for AI agents over MCP

Semurg: local persistent memory for AI agents over MCP

Comments
9 min read
loading...