Why Isn’t AI Making Your Engineering Process 10x Faster?, Prakash Sundaresan, Defang
From AI-assisted development to an AI-native software factory
Most engineering teams are already using AI tools like Claude, Cursor, and CodeRabbit. They make coding 10x faster - but the overall process of turning an idea into reliable production software often improves much less.
Why? Because writing code is only one step in the system. Requirements, design, review, testing, deployment, security, observability, and responding to production issues still depend heavily on people. Speed up one step enough, and the bottleneck simply moves somewhere else.
This session explores what it takes to rethink the entire software development process around AI - moving from isolated AI tools toward an AI-native software factory that can continuously move through:
Intent → Design → Build → Test → Deploy → Observe → Learn → Improve
The goal isn’t to remove humans. It’s to remove humans as the glue between disconnected tools and workflows, while letting founders and engineers decide where judgment, oversight, and approval add the most value.
What participants will get out of it
- Assess where AI is — and isn’t — accelerating your development process
- Identify the human bottlenecks preventing larger productivity gains
- Understand what an AI-native software factory looks like in practice
- Explore how AI agents can work across the full software lifecycle, not just coding and reviewing
- Leave with concrete ideas for what to automate next and how
The session will combine a short framing, a live AI-native maturity assessment, practical examples and demos, and an open discussion about what participants are trying in their own companies.