Moving legacy with AI | Michael Hladky @ AI India
About the event
Legacy migrations are slow, expensive, and error-prone — unless AI is used properly. In this talk, I’ll show how we reduced migration time by 75% using a structured AI workflow. Instead of experimenting with prompts, the focus is on reproducible processes: AI context and token strategies, MCP, orchestration files, and the systematic evaluation of runs.
The emphasis is on what others can practically adopt to modernize large legacy codebases faster, more safely, and in a measurable way.
Takeaways
Why context and token management determine the success or failure of AI-driven migrations
How MCP and orchestration turn AI into a reliable engineering tool
How to evaluate, compare, and improve AI runs
A field-tested AI workflow that drastically reduces migration effort
Concrete patterns that can be directly applied to your own legacy projects
Preview of the talk
Keep exploring
Deterministic AI Workflows for a Large-Scale Code Base | Michael Hladky @ ngLodz
ngLodz
May 20, 2026 · Lodz
Deterministic AI Workflows for a Large-Scale Code Base | Michael Hladky @ ngKato
ngKato
May 21, 2026 · Katowice
Deterministic AI Workflows for a Large-Scale Code Base | Michael Hladky @ ngPoznan
ngPoznan
May 18, 2026 · Poznan
Deterministic AI Workflows for a Large-Scale Code Base | Michael Hladky @ Angular Wroclaw
Angular Wroclaw
May 14, 2026 · Wroclaw
Keynote Talk | Michael Hladky @ NgIndia 2026
NgIndia
Apr 11, 2026
Debugging Angular SSR: Tales of Pending Task, Hydration, Race Conditions | Julian Jandl & Michael Hladky @ NgIndia 2026