Chris Parsons
- talk - Technical Track - Developer Experience
Beyond Autocomplete: Mastering AI Coding Tools for Real Production Work
Most developers have tried AI coding assistants, but few are using them effectively for real production code. This talk cuts through the "vibe coding" demonstrations to show you practical techniques for getting genuine productivity gains from tools like Cursor, Copilot, and Claude Code.
Drawing from experience shipping AI-assisted production code daily to tens of thousands of users, you'll learn battle-tested strategies including: the "reset and refine" approach that outperforms continuous iteration, why your AI is wrong 70% of the time (and why that's fine), and how to manage the unique cognitive load of AI pair programming.
We'll explore practical techniques like repository-specific rules that prevent repeated mistakes, using documentation as a collaboration tool between humans and AI, and the critical guardrails that make AI-generated code production-ready.
You'll see how to progress from basic autocomplete through to autonomous agents, with hands-on examples of context management, prompt stacking, and the documentation-driven development approach that transforms vague requirements into focused AI assistance.
This isn't theory - it's practical advice you can implement immediately to transform your development workflow while maintaining code quality and your sanity.
About Chris Parsons
Chris Parsons is an AI consultant and technology leader who ships AI-assisted production code daily and helps team navigate the AI jungle safely. He is also the CTO and co-founder of Cherrypick, where he's built production AI systems serving 500K+ users and raised £5M in VC funding.
With over 20 years of software development experience, he helps technical teams transform their development practices through practical AI adoption. His practical approach has helped trained teams double their output without experiencing the typical J-curve productivity drops.
Chris regularly shares his experiences building with AI through his blog, newsletter, and speaking engagements, focusing on the practical realities rather than the hype. He combines deep technical expertise with a pragmatic approach to help developers master the new world of AI-assisted development