Abhinandan Nuli

I’m a solution-focused AI Engineer with 7+ years across AI product engineering, data science, and applied statistics. I help teams translate real-world requirements into machine learning and LLM systems that are dependable, measurable, and production-ready.

In my most recent work, I architected an AI workflow orchestration system that converts natural-language requirements into executable JSON workflows. I designed deterministic execution with dependency-aware ordering and per-step logging, and built a semantic workflow retrieval layer with ChromaDB to eliminate hallucinated tool names.

I’ve also delivered end-to-end GenAI experiences across healthcare and enterprise use cases—building GPT-powered RAG systems with knowledge-base retrieval, chunking for token limits, and embedding-based search. Along the way, I benchmarked multiple LLMs across providers using evaluation ledgers to isolate prompt vs. model failures and improve step routing.

Earlier in my career, I led predictive modeling and segmentation efforts that improved business outcomes, and produced decision-support insights for stakeholders. I bring a strong statistics foundation—feature engineering, hypothesis testing, and forecasting—so my AI solutions aren’t just accurate, but actionable.