your agent's memory isn't slow because of the algorithm. its slow because of how the graph is represented.
Adjacency lists + zero-allocation traversal = real-time agents that actually keep up ๐
https://t.co/zurn0K745r
We scored 70. The competition scored 27.
We benchmarked Minns on StructMemeval.
Minns outperformed other systems across all categories, with especially strong results in multi-hop state reasoning.
Even when paired with a stronger model, the competition only reached 62%
AI infra is moving fast ๐ฎโ๐จ
RAG โ retrieve knowledge
Agentic RAG โ use tools
Memory โ agents that remember
We looking towards whats next, agents that learn, predict, and decide.
Hereโs how the stack is evolving
@s_chiriac We are building a memory/intelligents layer for agents:
@minns_ai
Its early days but its a fun jounrey, looking for real feedback at this point. If you have an agent, please feel free to plug it in and play around ๐ค
We used our @Nodescript_dev#nocode platform to automate meeting summaries and post them to Slack. Took about 10 mins to throw this together. Easily fits on our free plan. Hit me up in replies/DM if you want this for your team