The New Back Office: AI, expense reporting, and the Long Tail

There is a version of this story that is mostly hype. There is another version, the one we are interested in, that is mostly engineering.

Teams that win with computer-use agents tend to share a habit: they write the evals before they write the prompts. Everything else follows from that.

The skeptical read is that we are watching a feature, not a platform. The optimistic read is that computer use is exactly the kind of feature that becomes a platform when nobody is paying attention.

The cost curve matters here. Qwen 3 is roughly an order of magnitude cheaper per token than the equivalent model 18 months ago, and that changes which problems are worth automating at all.

The skeptical read is that we are watching a feature, not a platform. The optimistic read is that structured outputs is exactly the kind of feature that becomes a platform when nobody is paying attention.

Teams that win with RAG-as-a-service tend to share a habit: they write the evals before they write the prompts. Everything else follows from that.

Zendesk has been quietly running customer support through Granola for months. The results are unglamorous and, for that reason, more interesting than another benchmark chart.

None of this guarantees a clean story. Cohere could ship a model next month that rearranges the assumptions in this piece. But the direction of travel, for now, is clear enough to plan around.