AI Engineer / Agentic Systems / Multi-Tier RAG / Full-Stack Products
Built multi-tiered RAG & multi-agent systems with LangGraph, parallel debate nodes, state constraint schemas, and critique loops.
Engineered hybrid search pipelines (Pinecone + BM25) and Jina Reranking, optimizing latency (MTTV) by ~51%.
Designed non-generic premium client frontends using Next.js 15 & React 19, dockerized environments, and custom MCP integrations.
Each project is written for a fast scan first, then a deeper read: problem, approach, system, stack, and proof.
Messy public debates and complex legal scenarios resist simple, consensus-driven answers, and LLMs under prompt constraints frequently suffer from token overruns and perspective compliance issues.
Most AI assistants answer the current prompt but lose long-term builder context, project momentum, and evolving preferences.
Job discovery usually relies on brittle keyword matching, forcing candidates and recruiters through noisy search flows.
Startup platforms usually optimize visibility and pitch polish, but early ideas need critique, trust, milestone clarity, and emotionally-aware validation.
The portfolio is intentionally weighted toward what hiring teams can evaluate: tools, project choices, implementation tradeoffs, and communication.
When systems are built correctly, the interface gets out of the user's way and the underlying logic feels obvious.
A product lens changes ML architecture: the goal becomes useful agency, not only accuracy.
Good data products expose structure before decoration, especially when decisions depend on the signal.
I am available for AI/ML engineering opportunities where retrieval, agentic workflows, and product implementation meet. The fastest way to evaluate me is through the case studies and GitHub work.
