I build polished, production-minded AI/ML demos and infra - fast prototypes that scale, reproducible pipelines, and readable code that teams actually use. Expect clean demos, reproducible experiments, and practical infra patterns.

Portfolio • LinkedIn • GitHub • Email

• Multimodal demo: lightweight audio/image/text → app pipeline with low-latency inference and a friendly demo UI.
• Retrieval-augmented app: vector search + caching + reranking for accurate, fast semantic search or QA.
• Production ML infra templates: reproducible training loops, experiment tracking, CI for models, and monitoring dashboards.

These are concise, linkable repos I publish with clean READMEs and runnable examples.

• Languages: Python, TypeScript, SQL
• Core ML: PyTorch, TensorFlow, scikit-learn, XGBoost
• LLMs & Multimodal: Hugging Face Transformers, LangChain, RAG patterns, Whisper, Diffusers
• Data & Infra: PostgreSQL, MongoDB, Redis, Spark, Airflow
• Serving & Infra: FastAPI, Docker, Kubernetes, GCP, Azure
• MLOps & Observability: MLflow, Weights & Biases, Prometheus, Grafana
• Tooling: Git, GitHub, CI/CD, reproducible experiments

• Start small: prototype under an hour, then scale the winner.
• Tests for data contracts and model I/O before serving.
• Lightweight APIs and clear examples so non-ML engineers can integrate quickly.
• Focus on latency, cost, and reproducibility - not just score-chasing.

• Full-time roles in AI/ML engineering, ML infra, or senior prototyping
• Collabs on demos, tooling, projects, and reproducibility