Louis Knight-Webb combines technical brilliance with rare composure.As co-founder of bloop.ai and a Y Combinator S21 alum, Louis is building AI tools that don’t just touch legacy systems - they understand them. From mainframes to COBOL, his team is using generative models to translate decades-old infrastructure into clear, maintainable modern code.In this BASELINE episode, Louis breaks down:
– Why mainframes still power trillions in daily transactions
– How AI can turn legacy code into natural language
– The real blockers in enterprise modernisation (and how to overcome them)
– Why modernisation is about translation, not disruption
EPISODE CHAPTERS:
00:00 – Intro: Why legacy still matters
01:00 – Mainframes, batch jobs, and modern AI
03:00 – Louis on Y Combinator and the founder mindset
08:30 – What bloop actually does
14:00 – AI vs enterprise systems: reality check
21:00 – Turning COBOL into natural language
28:00 – Real-time systems, Kafka, and the death of the calendar
35:00 – Technical debt, team knowledge, and scaling modernisation
43:00 – Future-looking: language models, trust, and quantum
51:00 – Final reflections: calm innovation and quiet disruption
