🎥 Recorded live at the MLOps World | GenAI Summit 2025 — Austin, TX (October 9, 2025) Session Title: Agent-Powered Code Migration at Realtor.com Speaker: Naveen Reddy Kasturi, Staff Machine Learning Engineer, Realtor.com Abstract: Migrating large-scale ML and data workflows is notoriously tedious, error-prone, and difficult to automate. At Realtor.com, Naveen Reddy Kasturi led an ambitious migration of nearly 100 ML pipelines from a legacy self-managed Metaflow setup on AWS Batch to a fully managed Outerbounds Metaflow environment on AWS EKS—without breaking functionality or slowing innovation. In this lightning talk, Naveen shares how the team combined static code analysis, AI-powered pattern recognition, and LLM-assisted rewriting (using Claude via AWS Bedrock) to accelerate and scale the migration process. The result: 30,000+ lines of code migrated in weeks instead of months, using an AI-assisted workflow that generated blueprints, rewrote code, and enabled human-in-the-loop validation for accuracy and compliance. Beyond the migration itself, this session demonstrates how AI can act as a force multiplier for MLOps, turning painful, manual platform transitions into repeatable, inspectable, and scalable workflows. What you’ll learn: • Why effective AI-assisted migration depends on static analysis, pattern recognition, and domain-specific context • How to design scalable, AI-augmented code migration pipelines • Why human-in-the-loop validation is essential for correctness and compliance • How pattern recognition unlocks platform-level automation for future migrations • How generative AI is transforming developer workflows and infrastructure management in MLOps.