Скачать или смотреть Autonomous ML Agent MVP | Data Ingestion, Preprocessing, Leaderboard & Deployment

In this video, I walk through my Headstarter sprint project: an Autonomous ML Agent that automates key parts of the machine learning workflow for tabular datasets.

✨ What’s working so far:

✅ Project structure and dependencies set up

✅ Data ingestion and preprocessing pipeline (handle missing values, categorical encoding, datetime expansion)

✅ Leaderboard interface to compare candidate models

✅ Deployment artifacts and scripts (Python package / FastAPI service)

🚧 Coming next:

Model training with hyperparameter optimization

LLM-based orchestration for code generation and refinement

Meta-learning warm starts

Model ensembling strategies

This MVP already lays the foundation for an end-to-end pipeline that can be expanded into a fully autonomous ML system orchestrated by LLMs.

🔔 Subscribe to follow my journey as I iterate on this project sprint by sprint!