Charlie and the Software Factory & From Prompt to Loop

A double-header evening with two talks.

Charlie and the Software Factory
Everyone wants a golden ticket: a software factory that turns ideas into secure, production ready software on demand. The newest factories run themselves. AI agents plan the changes, write the code, review each other's work, and iterate, with hardly a human on the factory floor.

In this talk, we'll take the full tour. We'll see where the humans now stand. We'll meet the Oompa Loompas of the modern factory: the agents doing the work, and the paved roads, conventions, and written-down knowledge they depend on. We'll confront the Willy Wonka problem. And we'll visit the kids who didn't make it out: the common ways an AI-powered factory go wrong, and how to avoid becoming a cautionary tale.

Charlie Allen
Charlie has an unconventional background, moving from legal training into tech, teaching English in Korea, and co-founding a chai business before joining Amazon Alexa's AI organisation, where he deployed a production NER model. He was named the British Computer Society's Digital Degree Apprentice of the Year, later led infrastructure recovery at Darktrace, and went solo in 2026 to build agentic AI systems full-time from Norwich.

Charlie's LinkedIn

From Prompt to Loop
A practical guide to AI agent development, walking through four evolutionary stages: prompt engineering, context engineering, harness engineering, and loop engineering. Includes concrete examples using Claude Code and agent verification techniques.

Nishant Kumar
Nishant's LinkedIn