I’m a software developer and the founder of Nexusnext, currently focused on building next-generation knowledge infrastructure for AI agents.

My primary obsession right now is solving “context rot” in RAG pipelines and multi-agentic workflows. Standard semantic search is blind to time, so I spend my time building deterministic ranking systems and temporal decay engines to ensure LLMs only ingest mathematically fresh, high-velocity data.

We recently rolled this architecture into the Knowledge Universe API, allowing developers to filter out stale context before it hits the prompt, drastically reducing token bloat and eliminating hallucinations caused by outdated facts.

Always open to talking about custom LangChain retrievers, agentic system design, or fixing broken RAG architectures. Feel free to shoot me a DM if you want to test out the Knowledge Universe API endpoint in your own stack!