Designing a Production-Grade MLOps Pipeline with Open-Source Tools | Live Problem Solving Session #2

Hosted by: Siddhant & Toni Ramchandani

Bootcamp Exclusive

Sun, 01 Feb 2026
07:00 PM - 08:30 PM IST

Online (Zoom)

Hosted by

Siddhant

AI Research Engineer

Toni Ramchandani

Vice President

Share With Friends

Training a model is the easy part. The hard part is everything around it: reproducibility, data validation, experiment tracking, model promotion, serving, and monitoring for drift.

In this live session, we’ll build and run a complete open-source MLOps pipeline end-to-end, from data preparation to production serving and monitoring.

You’ll see the full workflow running as real code with a production-style project structure (package + config + tests + CLI), and you’ll leave with a working template you can reuse in your own projects.

This Session is Best for You if

✅ You build or deploy ML models and want a practical MLOps workflow

✅ You want to understand tracking, registries, and “Production model” promotion in practice

✅ You care about reproducibility, validation, and monitoring — not just training accuracy

✅ You want a real demo with working code you can run yourself

What We’ll Cover

• Production-ready ML project structure (package, config, tests, CLI)

• Data pipeline with train/val/test splits, saved artifacts, and lineage tracking

• Automated data validation and quality gates to block weak models

• Model training and experiment tracking with sklearn + MLflow

• Model registry workflow with Production promotion and FastAPI serving

• Monitoring in production using Prometheus metrics and drift detection (PSI + KS)

Pre-Requisites / Pre-Read

• Basic understanding of machine learning concepts and workflows

• Familiarity with Python and running ML code locally

• Interest in taking ML models beyond notebooks into production