Vercel shared its post-training harness with frontier AI model labs
Jared Palmer · ⚡ Inside GitHub’s AI Revolution: Jared Palmer Reveals Agent HQ & The Future of Coding Agents · Nov 10, 2025 · at 10:33
Jared Palmer, former VP of AI at Vercel and SVP at GitHub, describes Vercel's strategy to improve frontier LLM coding capabilities on Next.js.
0:00 / 0:22exact quote · 22.7s
▶ Watch the full episode on YouTube → 720p mp4 · rendered on demand · StarZero watermark
“We also started working with all the frontier model labs to help because it was in our Vercel's best interest. To have them be created Next.js. And also because of the post train models, and you can read about this on the Vercel blog, like we can the post training like harness that we created, we were started sharing and stuff with other model labs and stuff like that. And we had all our data in a very hygienic state to work with them.”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →
More from Jared Palmer
Vercel's v0 added $1M MRR every 14 days after chat rewrite
“When we launched V-Zero, the chat version, or the new V-Zero, whatever you want to call it it's like, 14 days, another million MRR, 14 days, another million MRR, it was like a rocket ship after that.”
Constraining v0 to Next.js and shadcn enabled faster execution
“What was really liberating for us was actually the focus on just one stack, or one framework, whatever else was trying to do, General purpose coding agent. We were like, no, we're just going to focus on Next.js front-end and Chad CNN. And that was really, that…”
Autonomous AI agents require vastly more reliable infrastructure to succeed
“Agents will really only work where we get to not only like the more intelligent models, but better reliability of the infrastructure providers, right?”
GitHub rejected stacked diff prototypes in 2020 and 2022 as too risky
“And there's been multiple attempts at this internally going back to like, 20, 20. And there was one very, very, very polished attempt to in 20, 22. And it just, I don't have all the context, so, but it was, there was a pretty good implementation. All of the wo…”