Production-style AI/ML engineering projects focused on model infrastructure, agentic RAG, evaluation, and research-to-production workflows.
System design
Service boundaries, data models, async jobs, and operational workflows.
ML evaluation
Benchmarks, metrics, reproducibility, and failure-mode visibility.
Research to production
Pipelines designed to move from experiment to reliable engineering artifact.
Each project is intentionally lightweight but structured like a real system: problem, architecture, implementation choices, status, and future work.
ML infrastructure platform for model registry, inference, evaluation tracking, async jobs, and benchmarking.
Agentic RAG backend for engineering knowledge, code, logs, tool calling, and grounded answers.
Research-to-production ML backtesting framework with leakage-safe validation and transaction-cost-aware evaluation.