mlengineering

✨ Build a machine learning model from a prompt

Cryptocurrency trading engineering: A scalable back testing infrastructure and a reliable, large-scale trading data pipeline

ML Engineering best practices | Deploying an ML API with FastAPI, PostgreSQL, and Docker for scalable model storage and inference.

ML Engineering best practices | ML model deployment as a FastAPI service, containerized with Docker for scalability and reproducibility.

End-to-end MLOps project demonstrating production ML engineering: automated data pipelines, multi-model training, MLflow tracking, model registry, FastAPI deployment, drift detection, DVC, Docker, and GitHub Actions CI/CD.

Enterprise reference architecture for Agentic AI systems using LangGraph, RAG, knowledge graphs, and LLM reasoning to power autonomous operational decision workflows.

This repository contains my implementations, experiments, and projects related to Machine Learning algorithms, models, and data science workflows.

Computer vision pipeline for pasture biomass prediction using DINOv3 embeddings, Ridge regression, reproducible inference, Pytest, and GitHub Actions.

It is easy to prototype ML models. With higher levels of abstraction, the expertise required to build any ML model is decreasing each day.

Config files for my GitHub profile.

Repository for project work of Udacity's Azure ML Engineer nanodegree.

Hand-implementing an N-BEATS model from the paper, to compare it against the classical methods (SARIMA and ETS) on large mortgage-related datasets

🔐 Production-grade credit card fraud detection system — XGBoost + SMOTE on 284k transactions (99.9% acc · 93.2% F1 · 98.7% AUC). FastAPI serving, Kubernetes deployment with HPA autoscaling, full MLOps stack: DVC · MLflow · Prefect · GitHub Actions CI/CD · 90%+ test coverage.