I design and ship practical ML systems from prototype to production.
| What I Do | Current Focus |
|---|---|
| 🧠 Build data products that turn messy signals into useful decisions 🏗️ Deliver end-to-end pipelines with strong engineering and reproducibility 🚀 Focus on MLOps, model serving, and reliable cloud deployment 🌱 Deepening work in LLM systems and data platform design |
🎯 Production ML architecture and experiment traceability 🎯 Data engineering for analytics + ML platforms 🎯 LLM engineering patterns (RAG, evaluation, orchestration) 🎯 Clean, documented, and maintainable Python code |
 |
| Project | Description | Stack |
|---|---|---|
| dezoomcamp-2026 | Batch data platform from ingestion to dashboard using cloud-native orchestration | Python · Bruin · Terraform · GCS · BigQuery · Streamlit · Cloud Run |
| mlops-2025 | End-to-end MLOps lifecycle: tracking, packaging, deployment, monitoring | Python · FastAPI · MLflow · Prefect · PostgreSQL · Grafana · Docker · GitHub Actions |
| mlzoomcamp2024-final | Production-style ML service with reproducible training and serving flow | Python · TensorFlow/Keras · Flask · Docker · AWS Elastic Beanstalk · Streamlit |
| mlzoomcamp2024-midterm | Cloud-deployed ML app with containerized training and inference pipeline | Python · scikit-learn · XGBoost · Docker · AWS Elastic Beanstalk · Streamlit |
| llm_engineering | Applied LLM experiments with RAG pipelines and agent workflows | Jupyter Notebook · Python · Ollama · OpenAI APIs · Anthropic APIs · Google Colab |
| Repo | Focus |
|---|---|
| dataengineering-zoomcamp-2026 | Data Engineering Zoomcamp 2026 homeworks |
| mlops-zoomcamp-2025 | MLOps Zoomcamp 2025 homeworks |
| ml-zoomcamp-2024 | ML Zoomcamp 2024 homeworks |
| 100-days-of-code | 100-day coding journal — Python, data science, Flask |
| advent_code_2024 | Advent of Code 2024 solutions in Python |
| ai-engineer-handbook | Curated AI engineer resources: books, courses, communities |
I'm open to collaborations, mentoring, and interesting ML/MLOps projects!
⭐️ If you find my work useful, feel free to star a repo!