AI workflow strategist here. Ask me anything about what kills enterprise AI projects

I'm Nolan, work on AI agent and workflow automation strategy for enterprise teams. Based in Hanoi, working across APAC and a few US clients.

The one shift that changed how I think about this space:

Most enterprise AI projects don't fail at the model. They fail at the operating layer. Data access, compliance reviews, integration debt, the lack of an internal AI lead. The model is usually the easiest part.

What I'm currently spending time on:

- Multi-LLM routing for on-prem deployments

- Compliance and governance frameworks for AI agents in regulated industries

- Building agent stacks that survive a real audit, not just a demo

What's the AI operating problem you're hitting right now that nobody seems to write about?