Blog
Thinking out loud
Case studies, engineering notes, and honest observations about AI in real-world operations.
From 15 minutes to 30 seconds: how an agent changed our order processing
A wholesale distributor processing 200+ orders a week was spending 15 minutes per order on manual ERP entry. Here is exactly how we got that to 30 seconds.
6 min readWhy decision intelligence beats RPA for operations teams
RPA breaks the moment your supplier changes their PDF template. Decision intelligence learns what you meant, not just what you did. Here is the difference.
Building AI agents that live inside your ERP, not alongside it
The biggest mistake in ops automation is building tools that sit next to your ERP. The agent has to be in the loop, not an afterthought. Here is how we architect for that.
The most important thing an agent does: handling exceptions well
Automating the easy 80% is trivial. The value — and the risk — is in the 20% of orders that are weird. Here is our approach to exception handling.
AI agents in manufacturing operations: what is actually working in 2026
A clear-eyed look at what AI is genuinely delivering in manufacturing floors and back offices today — and where the hype still outpaces reality.
The right way to start with AI agents: pick one workflow and go deep
Every successful deployment we have seen started narrow and went deep. Here is the framework for choosing your first workflow and measuring its success.