Tetriz Harness
There's a real problem hiding inside this pitch, and I've seen it play out across the teams I track. One person figures out a killer prompt for the support tickets, another cracks a workflow for code review, and the rest of the company keeps asking them how they did it. Institutional knowledge has always been sticky, but AI has made it worse. Now the expertise lives inside a chat window, and it doesn't transfer unless someone literally copy-pastes their setup.
Turning that tribal knowledge into an actual product is smart, because it hits a pain point every scaling company feels. The demos that get me excited are when I see a junior hire get the benefit of what the senior engineer already figured out, without a boring training session.
What I'd want to know is how they handle the mess. Team learning is unruly, and someone has to decide what's worth promoting to the shared setup. I'd also push on whether they curate for quality or just let everything pile up. If the harness becomes a dumping ground, it loses its magic. Still, the instinct to productize something this granular is solid, and I'm curious to see whether teams actually share their best configurations or guard them like trade secrets.