a24z - Agents with Superpowers: Context Engineering Hackathon
AI Tinkerers - San Francisco
Hackathon Showcase

a24z

Team led by a24z founder Brandon In — hands-on TypeScript/JavaScript AI engineer (a24z-memory, context engineering), active in SF hackathons (Sept–Oct 2025).

1 member

We are using Redis with LangCache, Vectors, multi-agent orchestration.

We are creating a tool that allows agents to self-evolve. It takes in all of your own data from your AI Coding Agents and provides meaningful improvements so that you can reduce your token cost, improve your accuracy, and reduce frustration.

We utilized Redis langcache to cache our prompts and outputs, we used vectors in order to semantically search against all of our subagents that we created to determine which would be the best to provide to our customers based on their own repository and session data.

We used NextJS with OpenAI with Agents that talk to each other to accomplish a self-evolution..

I build an observability tool as a company, but I created an agentic solution that allows us to actually evolve based on our data.

AI Tinkerers Composio Github OpenAI Redis