Software Factories: A Field Conversation when code moves at machine velocity

Software development is becoming one of the clearest early examples of what happens when AI moves from assistance to delegated execution.

In this Summit Evidence AI Field Conversation, I sit down with TensorLake founder Diptanu Gon Choudhury to examine how the engineering bottleneck is changing: developers are beginning to manage parallel agent work, architecture and testing are becoming more important, and the traditional SDLC is starting to look less like a linear handoff and more like an objective-led operating loop.

The broader enterprise implication is not simply faster coding. Organizations need a better operating model for machine-generated work: clearer objectives, stronger evidence, risk gates, runtime infrastructure, and feedback loops that can keep pace.

Watch the full conversation or use the audio player below.

Disclosure: I advise TensorLake. This is a Summit Evidence AI field conversation about software development, AI infrastructure, and enterprise adoption. It is not an independent product review.