Perla Gámez

You will meet a lot of amazing founders But you won't meet someone that loves her start up as much as I

Hiring genzers was the best decision I've ever made

I love getting texts like this

These 6 principles learned through trial and error will get you past the early stages of product development - @david.garnitz/product-principles-from-an-early-stage-ai-startup-ae0be130d36f?postPublishedType=initial" target="_blank" rel="nofollow noopener">medium.com/@david.garnitz…

“Why don’t you have more rooms?” Said my niece to me today while visiting my 1bd in SF

Super happy to see @UCBerkeley evolving their curriculum to adapt to the new world!

SOMEONE MADE A DIGITAL WHIP TO MAKE CLAUDE WORK FASTER 💀

7 security incidents in 7 days. npm. LiteLLM. Claude Code. Axios. Mercor. Delve. Railway. My cofounder Upal Saha and I wrote something about what's happening to this industry. The short version: laziness is destroying the software industry, and powerful tools are making it worse because nobody's checking the output.

They say your phone knows more about you than your mom. So why can't Siri tell me things like how much I spent on food delivery this month? DM me if you want to try a phone that can.

LLMs are the hardest workers you will ever meet

@BryanOnel86 Bryan, I don’t benefit from being dishonest here.

That is pretty dishonest of you Perla, to publicly call our company out on something that went down very differently. We evaluated your docs as part of our internal security DD process after YOU physically approached us at SaaStr, then reached out to get us on a demo call, after which we evaluated your product. Our DD process led to the conclusion that your security was insufficient, as a result of having boilerplate documents that contained "ACME" and having policies of which the placeholders were never filled out. Why would you lie about this? Smh

@LiuLawrence45 @BryanOnel86 @LiuLawrence45 We’ve migrated to Vanta. But I once had an engineer from OneLeet who faked interest in foam just to get access to our reports from Delve.

Dude in all honesty, it’s been ~5-6 days of you just straight dunking on Delve. It was funny in the beginning, but your Twitter has become a massive reason why we’re opting in to move to Vanta instead of Oneleet (we had heard amazing things about you guys before). Why feel the need to gloat / continually put them down, genuine question? You’ve won, do you not have more important things to do as CEO?

What did I just watch

Applications for S26 of the @thehousefund AI Accelerator are now open 🚀 We back founders building AI-first companies that embed AI into the startup’s DNA—not just add it on. Small cohort. Hands-on partners. Real progress. Apply now → thehouse.fund/ai-accelerator Go Bears 🐻

I need to seriously step it up

@abustamante 🐇

Happy Bad Bunny day y'all 😎🐰

I have been putting off showing our foam.ai site some love Spend some time today tweaking it Most of our work goes into quiet literal "backend" so it feels good to get an improvement that is super visual

arxiv.org/abs/2507.16075: “Deep Researcher with Test-Time Diffusion” Loved it for being niche and a useful mover for our performance.

arxiv.org/html/2505.0812…: “Putting It All into Context: Simplifying Agents with LCLMs” Loved it for its help on keeping foam scalable.

Last three months of 2025, I have been getting into the habit of reading arXiv subscription results and what foam engineers share, like I am a priest and arXiv and what they share contain religious scripts :) Here are my top 3 reads

My favourite paper of the year — LLMs are already having a large impact on the number of scientific papers produced!!