Startup Speed in Enterprise Code

Are you jealous of all these cool startups punching out new products and features at the speed of light and you’re stuck artisanally crafting features for a 5+ year old product? I have been working on large scale enterprise software for over a decade. The challenges faced in larger enterprises are very different to early-stage startups and it can be difficult to integrate AI tooling and generate trust in the outcomes. I will demonstrate my approach taken to integrate autonomous coding agents with larger, scarier code bases which have difficulty fitting in your context windows. The approach involves required pipeline automation for large projects, harness setup and using AST parsing rules to enforce architectural constraints reliably so you can spend less time going back and forth with a chatbot and more time getting value out to production. AST parsing is now easily accessible to all, it used to be a specialist domain but given a high-level understanding, you can now easily build your own powerful rules to guide your agents with compile-time feedback.