Uses

The tools I actually reach for. Not a skills list, just what's open on my machine most days. Tools change every year. The math, statistics, and systems thinking underneath them don't.

Languages

  • Pythoneverything ML/data. First language I think in.
  • SwiftiOS projects. Learning to love it.
  • SQLstill the most useful language I know.
  • JavaSpring Boot for backend services at work.

ML / AI

  • PyTorchmodel training and experimentation.
  • HuggingFacefine-tuning, embeddings, evaluation.
  • LangChainRAG pipelines. Useful until it isn't.
  • vLLMLLM serving. Fast and practical.
  • Gurobioptimization. Nothing else comes close for MIP.

Infra & Cloud

  • GCPprimary cloud. Vertex AI, BigQuery, GKE, Cloud Run.
  • Dockereverything gets containerized.
  • KubernetesGKE for production workloads.
  • FastAPIdefault API framework. Clean and fast.

Data

  • PostgreSQLfirst choice for structured data.
  • Qdrantvector search in production.
  • Snowflakeanalytics warehouse.
  • BigQuerywhen you're already on GCP and need scale.

Apps & Tools

  • Claude Codedaily driver for coding. Changed how I work.
  • Cursorfor Swift/iOS work specifically.
  • Obsidiannotes and thinking.
  • Linearproject tracking for personal projects.

Hardware

  • MacBook Air M1does everything without complaining.
  • iPhone 16 Pro Maxalso my primary test device for sarvis.