Before joining venture capital, Scott Sage worked as a nanny, in an antique shop, in music A&R, and as a MATLAB researcher, and he now looks for technical founders who build strong systems while staying humble enough to let their team shine, gets involved early in helping startups find their first customers and message their product clearly to buyers, and has said that India's deeptech and AI scene may be one of the most underpriced areas in venture right now. He's also willing to break from his usual seed-stage focus when he feels strongly about a company, as he did with Onfido after already advising them, and he tends to look for hard infrastructure and integration problems, like the one he backed at TriggerMesh, that other investors might see as too technical to bother with.