Context was always the 10x engineer
The hard part was getting three companies’ knowledge in one place.
I think context explains most of what we call a “10x engineer.” They know the codebase, the business, and how things connect across teams. Coding skill matters, but knowing where to look often matters more.
Yesterday, we fixed a payment bug that had been around for two years. Once AI had access to the code and logs from all three companies involved—payments, backend, and frontend—the bug was fixed in minutes.
AI helped. But the biggest change was that the context was finally in one place. I think we could have had the same result by getting the right people from all three companies into the same room.
That was the hard part: support tickets, company boundaries, and finding time at short notice. The fix had always been easy. Assembling the knowledge wasn’t.
For SaaS, this means finding a way to bring your product’s context, and the AI that understands it, into the room where the problem is being solved.
Getting knowledgeable people together quickly could still be as big a win as AI itself. The bottleneck was organizational. AI can help bring that knowledge together, but only if agents get access to the context and a shared place to collaborate.
In our case, the agents communicated by email, with a human relaying messages between them. It worked fine. But we should build and demand better ways for agents to work together across companies.
Robin Westerlundh
← back to the index