Scuttlebutt

Scuttlebutt

A top AI lab pays $240K for build analysis instead of building it — Hermetiq CEO

7 takeaways: why Jake says builds, not code, now slow software teams, his plan to do for Bazel what Databricks did for Spark, a $25B market, $2.2M raised and Jason Calacanis's advice

Scuttlebutt's avatar
Scuttlebutt
Oct 31, 2026
∙ Paid

This Week in Startups · Oct 30, 2026

With Jake Newfield, Co-Founder and CEO of Hermetiq

A top AI lab pays $240K for build analysis instead of building it — Hermetiq CEO

Photo via hermetiq.com

One of the top AI labs is paying Hermetiq $240,000 to analyze the more than 2 million software builds it runs every day. Jake Newfield, co-founder and CEO of Hermetiq, made that the center of his pitch on This Week in Startups, in the LAUNCH Accelerator's weekly investor session.

Investors worry that the AI labs will build whatever a startup sells and wipe it out. Jake's evidence runs the other way: a lab that could have built his product itself pays for it every month instead.

"Because of Agentic Coding, tools like Claude, Cursor, and Codex, coding and creating code is no longer the bottleneck that it was a couple years ago for companies. The bottleneck is now the build process which is the process of compiling, verifying and testing code." — Jake Newfield

Jake is in LAUNCH's 37th cohort and had won its pitch competition two weeks running. Before Hermetiq he sold software at Salesforce and Cloudera and managed products at Akamai.

The full segment is covered here so you can skip it.

Here are the 7 takeaways that matter.


This Scuttlebutt Summary is available in full to premium subscribers. If you're not one already, please consider subscribing to keep reading.


Key Takeaways

  • Coding is no longer the bottleneck for software teams; compiling and testing code is, Jake says

  • A top AI lab pays Hermetiq $240K to analyze over 2 million builds a day

  • Hermetiq wants to do to Bazel what Databricks did to Apache Spark, working at the protocol level

  • Big companies buy infrastructure they could build, as AI labs buy Databricks and Okta, to keep engineers on their core product

  • A lab paying instead of copying is Jake's answer to VCs' moat worries

    • Jake says an OpenAI or Anthropic could build what a startup has overnight

  • Jason Calacanis's advice was to lead with the lighthouse customer and get a written testimonial

1. Builds Are the Bottleneck

Jake opened his two-minute pitch with the problem Hermetiq solves. Agentic coding tools write code fast enough that writing it no longer holds a company back. The slow step is now the build: compiling the code, verifying it and running its tests.

Jake says the build problem runs from the biggest AI labs to small startups

"Every company from the AI labs down to AI-native startups — that is becoming the bottleneck for their software development process." — Jake Newfield

Hermetiq's answer is Barn Owl, which Jason Calacanis, who hosts This Week in Startups, introduced as an AI build engineer.

2. An AI Lab Pays $240K

User's avatar

Continue reading this post for free, courtesy of Scuttlebutt.

Or purchase a paid subscription.
© 2026 The Golden Age LLC · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture