Over the past year, I've been developing intensively with AI, ADEs, and software agents, and I've become increasingly convinced that the main challenge is no longer generating code.
The real challenge is maintaining control over the architecture, design decisions, requirements, and context that guide AI throughout the entire development lifecycle.
Lately, I've been exploring how context governance, traceability, and continuous alignment between design and implementation can help build AI-assisted solutions in a more predictable and sustainable way.
What interests me most is how software engineering will evolve in this new era, where productivity is no longer the primary challenge, but rather the ability to maintain coherence, control, and quality as systems grow.
