“What the working model that we've used for all of our internal agent development is basically a tripod. It's a GTM engineer, a data scientist, and then the best person, the best subject matter expert in that function. And they sit down shoulder to shoulder and document the best practice for that particular function and then encode that in workflows that ultimately form an agent. We then keep a human in the loop so that human will basically QA the agent. The agent won't be allowed to autonomously execute its action. And then over time as the subject matter expert has less feedback for the agent, we finally pull that person.”
Jeanne DeWitt Grosser, Chief Operating Officer. Runs operations at Vercel and created the go-to-market engineering team behind its internal agents, so she is in a position to know their cost and staffing; she speaks as an executive of a company that sells AI tooling.
Vercel sells AI or agent tools or services, so it has a reason to publish flattering numbers about AI. The passage stays in the archive; read it with that in mind.
How this was checked
Checked against the source by a first reader, then confirmed independently by a second reader who fetched the source without seeing the first reader's notes (Independent review agent (Claude Sonnet), blind second pass from passage text and citation only, no repository access, 2026-11-08).
Whisper matches the whole passage word for word near 03:24, ignoring punctuation.
How to cite this passage
Jeanne DeWitt Grosser (Vercel), “Going from 10 to 1 SDRs. How We Deployed AI Across our Sales Org with Vercel's COO”, SaaStr AI (YouTube); talk at SaaStr AI Annual 2026, 2026-11-04, at 03:24. https://www.youtube.com/watch?v=wXHLbr4W-8s Archived at https://the-ai-record.com/passages/vercel-tripod-engineer-data-scientist-expert-human-in-loop-2026
Context. Vercel's COO, making the case that agents change the build-versus-buy decision, gives the running cost of the lead qualification agent she described at the start.