ML Engineer building production AI systems — RAG pipelines, MLOps, and predictive modeling.
I design and ship end-to-end AI systems — from data pipelines and model training through to deployment and monitoring. My focus is on building things that actually work in production, not just in notebooks.
- LLM & RAG Systems — retrieval-augmented generation, semantic search, LangChain, LlamaIndex, FAISS, vector databases
- MLOps & Deployment — model serving, pipeline automation, cloud deployment (AWS/GCP), Docker, experiment tracking
- Predictive Modeling — time-series forecasting, anomaly detection, classification and regression at scale
- Data Engineering — automated ETL pipelines, data transformation, business intelligence workflows
Languages: Python, SQL
ML/AI: Scikit-learn, XGBoost, LightGBM, CatBoost, TensorFlow, PyTorch, HuggingFace Transformers
LLM Tooling: LangChain, LlamaIndex, FAISS, ChromaDB, OpenAI API
MLOps: Docker, MLflow, FastAPI, GitHub Actions
Cloud: AWS (S3, EC2, Lambda), GCP
Data: Pandas, NumPy, Spark (basics), PostgreSQL
AI-powered job matching system using retrieval-augmented generation. Ingests job descriptions, embeds them into a vector store, and matches candidates based on semantic similarity.
LangChain FAISS OpenAI FastAPI
End-to-end time-series forecasting system for financial market data. Includes walk-forward validation, SARIMA, Prophet, and ensemble modeling.
Prophet LightGBM Time-series Python
Production-grade ML pipeline for surplus produce recommendation. 5-seed hybrid ensemble (LightGBM + CatBoost) achieving public AUC of 0.945.
LightGBM CatBoost Ensemble MLOps
Modular data workflow system for extract → transform → load operations. Designed for business intelligence and data sync use cases.
Python Automation Data Engineering
- 🔭 Building production-grade AI systems for real-world deployment
- 🌍 Competing on Zindi — applied ML challenges
- 💼 Available for roles in ML Engineering, AI Engineering, and Data Science
"I build AI systems that ship."