Technical Stack

My Technical Stack

A deep dive into the tools, languages, and platforms I work with on a daily basis.

Core Languages & Runtimes

  • Python: Primary language for ML and backend services. Knowledge scientific computing (NumPy, SciPy), Type Hinting (Pydantinc) and in AsyncIO.
  • SQL: Expert-level proficiency in complex window functions, CTEs, and query optimization for high-volume data.
  • Bash: Shell scripting for automation, system administration, and CI/CD entry points.

Machine Learning & Data Stack

  • PyTorch : Deep learning framework of choice. Experience building and deploying models, fine-tuning, distributed training, and scalable pipelines.
  • Transformers : Working with HuggingFace for NLP, Retrieval-Augmented Generation (RAG), and fine-tuning LLMs.
  • Model Optimization: Proficiency in ONNX, TensorRT, or quantization to reduce inference latency in production.
  • Vector Databases: Implementing and managing ChromeDB, and Pinecone for high-dimensional similarity search.
  • MLflow: Model tracking, experiment versioning, and registry management.

Platform & Infrastructure

  • Kubernetes (K8s) : Container orchestration, deployments, services, and monitoring. Familiar with Kubeflow for specialized ML workflows.
  • Docker : Containerization for reproducible environments and microservices.
  • Cloud (AWS/GCP):
    • AWS : SageMaer (Pipelines/Endpoints),S3, EC2, Lambda, Athena.
    • GCP : Vertex AI, BigQuery Cloud Run, GCS.
  • Terraform : Infrastructure as Code (IaC) for modular and reproducible resource managment.

Data Engineering

  • Apache Spark : Distributed data processing at scale.
  • Apache Kafka : Real-time event streaming and data pipelines.
  • Pandas / Polars : Advanced data manipulation, utilizing Polars for performance-critical, multi-threaded operations.
  • MLflow : Model tracking, versioning, registry, and serving.
  • Airflow : Workflow orchestration and data pipeline scheduling.

Databases & Orchestration

  • Snowflake/BigQuery : Cloud data warehouse for ELT pipelines, large scale analytics, and feature stores.
  • NoSQL : Redis for real-time features. DynamoDB for metadata.

DevOps & CI/CD

  • Testing: Comprehensive unit and integration testing using Pytest.
  • CI/CD: Automation via GitHub Actions or GitLab CI, including automated model testing and deployment.
  • Prometheus & Grafana : Monitoring and observability.

Development Tools

  • Code Quality: Strict adherence to Ruff, Black, and Mypy for maintainable, production-ready codebases
  • Git : Version control, branching strategies, and collaborative workflows.

Last updated: 2026. Always learning and exploring new tools that solve real problems.