Drew Bredvick (@DBredvick) / X

Drew Bredvick

4,944 posts

Drew Bredvick

@DBredvick

Growing GTM engineering (applied ai) ▲ @Vercel

Austin, TX

Joined May 2011

Replies Articles Media

Pinned
Drew Bredvick
@DBredvick
Dec 2, 2025
Hi, I'm the engineer that built out Vercel's first AI SDR agent mentioned below👋 AMA
Lenny Rachitsky
@lennysan
Dec 1, 2025
My biggest learnings from Jeanne DeWitt Grosser (ex-Chief Business Officer at @stripe, now @vercel COO): 1. What failed seven years ago now works with AI. In 2017, Jeanne tried to build a system at Stripe that would automatically personalize outbound emails based on company
Drew Bredvick
@DBredvick
Aug 7
Well at least _something_ is reading the full text of the bills now. Though we might indirectly end up governed by Claude?
nardos mengesha
@nardosamengesha
Aug 6
cali not being first on here is freaking hilarious sf getting tech mogged by dc
Drew Bredvick
@DBredvick
Aug 5
This is a crazy impressive result and you can try it out yourself: curl -fsSL app.primeintellect.ai/prime-agent/in… | sh
Prime Intellect
@PrimeIntellect
Aug 5
Introducing Prime Agent: A self-improving RLM harness for coding and long-running autonomous tasks. Designed to be both token-efficient and expressive through programmatic tool calling, context as a variable, multi-agent messaging, and a self-modifiable harness state.
00:00
Readers added context
Prime Agent's ARC-AGI-3 score of 95.5% is on the public evaluation set, which is susceptible to overfitting; official validation requires agreement with semi-private/private sets. Its RLM currently has recursion depth of 1, not deeper recursion. arcprize.org/policy primeintellect.ai/blog/rlm
Drew Bredvick
@DBredvick
Aug 5
We will look back at this time period and note that all the people calling for the bubble to pop were right, they were just 2-10 years too early. Your instincts are right, your timing is wrong.
Drew Bredvick
@DBredvick
Aug 4
note to self, don't forget to actually place the bet drew.tech/newsletter/las…