Deep-Dive Showcases

Full-stack SWE meets agentic AI — polyglot microservices on EKS, React/Next.js/Angular frontends, TypeScript migrations, and production agent harnesses with evals and RL alignment.

TensorRT + CUDA pipelines on Jetson Nano, concurrent C++ with OpenMP, and real profiling with NVIDIA NSight. Systems work across embedded hardware and GPU inference.

Architecting scalable backend systems, robust cloud infrastructures, and interactive frontend experiences.

Serverless AWS pipelines (S3, Lambda, Athena, Redshift), distributed Spark & Hadoop clusters across 60+ Linux nodes, containerized vector DB & RAG on on-prem GPU servers, and automated Linux systems administration at scale.

End-to-end ML in production — feature pipelines, model training & MLflow tracking, hybrid RAG retrieval, eval harnesses with LLM-as-judge, containerized serving, and drift monitoring.

Transforming raw data into actionable insights to improve healthcare outcomes, ensure compliance, and build enterprise AI infrastructures.

Designing agent harnesses, sandboxed execution environments, and deterministic eval pipelines for AI behavior.

Bridging the gap between intelligent algorithms and physical execution through edge AI and embedded systems.

Locating and classifying entities in dynamic environments, with real-time inference optimization.