
I build stuff @hellyeah_ai @finalround_ai @worktrialai Previously ML at Pinterest, Meta, AMD DJing as https://nitter.cf/t.co/Nczgmlz8nZ
San Francisco
Joined May 2012
- Tweets685
- Following159
- Followers556
- Likes690
Pinned Tweet
You don't need to be in SF to build a generational company.
NYC startups are cooking right now. A bunch just raised monster rounds and are hiring ASAP.
I did some digging and found ~15 companies with verified recent rounds and open roles.
Core-cluster offices are pinned; more that also raised and are hiring are listed alongside.
A few that jumped out:
→ Ramp ($1.85B)
→ Kalshi ($1B)
→ Modal ($750M Series D)
→ Cyera, EliseAI, Profound, Valon
Manhattan is where a lot of top tier talent is going to build right now.
If you are looking for a seat in NYC, this is the board.
We are hiring at Hellyeah AI.
jobs.ashbyhq.com/hellyeah/67…
Vanta crossed $300M ARR earlier this year.
I mapped the GTM org.
There are ~1,700 people currently at Vanta.
724 of them joined in 2026 alone.
Team by team:
1) Account Executives (~164)
→ 94 joined in 2026
→ Top AE sources: HubSpot (18), Salesforce (14), Kandji (5), CrowdStrike (4)
2) SDRs / BDRs (~110)
3) Customer Success (~119)
→ 57 joined in 2026
4) Solutions / Sales Engineers (~44)
→ 23 joined in 2026
5) Partnerships / Alliances (~38)
6) GTM / Product / Demand Marketing (~88)
7) Sales leadership (~13)
→ Kelly Bray: SVP, Global Sales (ex-MongoDB)
→ Eric Martin: VP of Sales
→ Julie Martin: VP of Sales, Americas (ex-Airtable)
→ Chris Sander: VP Sales EMEA (ex-Faire)
→ Sam Shoolman: Director of Sales, Growth (ex-HubSpot)
→ Willie Pierson: Global Head of Sales Development (ex-Gong)
Vanta appears to be hiring AEs who've already sold security and SaaS platforms, then promoting from inside for leadership.
Who should I map next?
3
10
365
Riot Games has become one of gaming's most successful companies.
Many employees leave to found startups.
Here are 12 founders from 6 companies started by ex-Riot Games employees:
1. Theorycraft Games: Joe Tung + Areeb Pirani ($87.5M raised across Series A and B)
2. The Believer Company: Michael Chow + Shankar Gupta-Harrison ($55M Series A)
3. Singularity 6: Anthony Leung + Aidan Karabaich (about $49M raised, then acquired by Daybreak in 2024)
4. Odyssey Interactive: Richard Henkel + Dax Andrus + 🦆 Eric Lawless (about $25M raised, incl. a $19M Series A)
5. Ruckus Games: Jesse McIntyre ($19M Series A led by Krafton)
6. Exia Labs: Jonathan Pan + Serj Kazar ($2.5M pre-seed)
3 patterns stood out:
1. Leaders raise the biggest rounds. Theorycraft, Believer and Singularity 6 raised about $190M combined. Their founders include Riot's former EVP of League of Legends and a Riot VP.
2. Not everyone stayed in games. Exia Labs, started by two ex-Riot Games colleagues, is building AI for military operations.
3. It's going global. I found 49 people with Riot Games in their work history who've started companies since 2024. 21 of them are outside the US.
Shipping a live game to millions of players is the best founder training there is. Riot Games alumni keep raising because they've already run the hardest version of a product.
1
10
416
Ramp's #1 source of hires isn't another fintech startup. It isn't Google or Meta either.
It's Oracle. 26 hires, and 20 of them came from NetSuite.
I pulled every LinkedIn-visible move in and out of Ramp since November 2024 with Crustdata (YC F24).
Ramp hired 1,010 people and lost 512 to new jobs.
Who Ramp hires:
→ 220 in sales (AEs, account managers, new business)
→ 188 engineers
→ 146 in customer success, support and implementation
→ 41 with AI or agents in their title
Where they come from:
→ Oracle: 26 (20 from NetSuite)
→ Shopify: 16
→ Stripe: 13
→ Verkada: 10
→ BILL: 10
Where they go:
→ OpenAI took 21, nearly 3x anyone else
→ Mercury, Clay and Stripe took 8 each
Big Tech sent Ramp 83 people. Ramp sent Big Tech 11.
Customer success: 146 in, only 30 out. Ramp is staffing up the people who keep customers, not just the people who close them.
My read is that Ramp hires the people who already sold and set up the finance tools its customers run on. Then its builders leave for AI labs and startups, not Big Tech.
The best sales hire is the one who already knows your customer's stack.
1
15
606
Being a Founding Engineer at the next OpenAI or Anthropic is one of most underrated ways to become a deca-millionaire.
The key is finding the right startups before they're massive.
I pulled live Founding Engineer (and founding-adjacent eng) postings at recently funded startups:
1. Raindrop ($35M Series A): Founding Engineer (Backend)
jobs.ashbyhq.com/raindrop/3d…
2. Prime Intellect ($130M Series A): Member of Technical Staff (Training Platform) ($150K-$300K + equity)
jobs.ashbyhq.com/primeintell…
3. EliseAI ($350M, stage undisclosed): Founding Software Engineer ($240K-$300K + equity + benefits)
jobs.ashbyhq.com/eliseai/6b7…
4. Norm Ai ($120M Series C): Founding AI Engineer
linkedin.com/jobs/view/43426…
5. Valon ($150M Series D): Founding Forward Deployed Engineer
linkedin.com/jobs/view/43665…
6. Higgsfield AI ($400M Series B): Founding Data Engineer
jobs.ashbyhq.com/higgsfielda…
7. Numeral ($100M Series C): Founding Solutions Engineer
linkedin.com/jobs/view/43185…
8. Serval ($150M Series C): Founding IT Engineer ($150K-$250K + equity)
jobs.ashbyhq.com/serval/a44f…
These companies just took capital. They're still small enough that a founding eng hire moves the needle.
Hellyeah AI is hiring this shape of engineer too (closest open role: AI Engineer: Learn Engine ($200K-$1M + 0.5%-3% equity)).
jobs.ashbyhq.com/hellyeah/67…
3
66
9,672
Jay Ma retweeted
Jensen Huang resigned as CEO of NVIDIA last Thursday.
When I heard the news, I sent him a $1M job offer.
He never replied.
A one-year-old AI lab with a couple dozen people just raised $1B at a $6B valuation.
Flapping Airplanes is the latest SF-based AI lab to raise a monster Series A, on the heels of a $180M seed earlier this year from GV, Sequoia and Index.
The founders are cracked:
- Ben Spector, CEO. Stanford CS PhD student. Founded Prod, the student-run accelerator, at MIT.
- Asher Spector. Stanford statistics PhD, former debate champion.
- Aidan Smith. Thiel Fellow, former founder, 3 years at Neuralink while in college at Georgia Tech.
Their team is extremely talented and consists of IMO, IOI and IPhO medalists.
One engineer joined while still in high school and built the agent infra that now automates much of their work. Another had never trained a model, then led their entire pre-training run.
Small, high-slope, high agency teams are the future.
These are the same kinds of people we're hiring at Hellyeah AI.
Who's the most cracked young researcher you know?
2
1
299
YC’s top 100 never turn ads off
44% of live YC companies run paid ads.
Adside checked all 6,055 against Google, Meta, and LinkedIn. Newer batches advertise less — the 2026 batch sits at 19%.
I checked F26. These AI apps are already buying growth:
@Kairos Health - @Tanush Chintala, ~69 active Meta ads
@Bayworks - @Kaustubh Bansal, ~37 active Meta ads
@antimattr - @Sridipto Ghosh, ~35 active Meta ads
@Adside.ai - @Robin Choy, ~30 active Meta ads
@Veeza AI - @Aly Moursy, ~18 active Meta ads
@Grep AI - @AJ Asver, 40 Google ads
@Agent37 - @Vishnu Krishnaprasad, 16 Google ads
@River - @Chandler Supple, 15 Google ads
How we counted: Adside = live ads across Google / Meta / LinkedIn. F26 lines = Meta active on 11/4, or Google ads for that domain any time. Neither is spend.
Study in the first comment.
F26 founders: is paid already working, or still a test?
9
21
628
what a great time to build - cookin something great :) stay tuned -
we actively hiring at hellyeahai.com/careers if you
- can ADHD maxxing and manage more than 10 agentic sessions at same time without losing track
- fascinate about scaling human with compute
- believe that THERE IS ALWAYS A WAY
- constantly hitting your cc / codex plan limit
- have crazy github stats
- CURIOUS AS FUCK
we should talk or cowork
we offer 1000$ token credits per day for anyone who cowork with us at our SF HQ
sf coworking includes $1000 in token credits per day. if you want to build with us, check hellyeahai.com/careers
1
291
We're hiring an AI engineer at Hellyeah AI right now for up to $1M TC.
I constantly get asked if this is legit or a scam.
The second most common question is - how will you even know when you've found the right hire? Super valid.
There are many people out there who are good at faking competence. Maybe they came from Big Tech and have a lot of traditional pedigree.
We're looking for 10x engineers, not 0.1x larpers who've never shipped anything in their life.
Pair code with me once and I'll know if you're 10x or 0.1x.
I only care about what you've shipped before, not what's on your resume.
Our process for vetting engineering talent is 3 steps:
1. Show me your GitHub.
2. Walk me through one project you actually built.
3. Pair code with me.
One session and I can literally tell if you can build or if you're just talking.
What gives it away:
- Do they pay attention to the technical design or just vibe it?
- Are they on the best tools? Do they know what dropped on X and GitHub this week?
- Can they context switch between humans and agents without dropping a thread?
- If they ask thoughtful questions
So yes we are paying up to $1M for the 10x to 100x engineers who can run dozens of humans and agents at the same time.
If this sounds like you, DM me your GitHub or apply on our site: hellyeahai.com/careers
Music for ALLLL
ACE Music v1 just hit #1 on @ArtificialAnlys' music leaderboard. It's the first model on this chart to outrank Suno's latest release.
Try it free at acemusic.ai.
We'll open-source the model weights in the next few weeks. We're making the final tweaks so you can run it locally.
1
189
Anthropic's IPO is set to mint thousands of new millionaires. One of the most underrated ways to build generational wealth is being a founding engineer.
If you want to be the next, you need to keep tabs on which startups are growing fast.
A ton of SF startups just raised and are hiring right now.
I mapped ~15 of them.
Street-address offices in the core cluster are pinned; more that also raised and are hiring are listed alongside.
The typical pattern involves a talent dense team raising money, then posting a bunch of open roles as they race to get top talent.
A few that jumped out:
- Cognition AI ($2B Series E)
- Harvey ($550M Series H)
- Databricks ($5B round)
- Together AI, Bedrock Robotics, Serval
If you are hunting for a Founding Engineer seat in SF, start here.
Btw - we are hiring a founding engineer at Hellyeah AI too.
jobs.ashbyhq.com/hellyeah/67…
Want to work at OpenAI? Start your career at one of their talent farms - Big Tech.
OpenAI hired 2,716 engineers in two years and lost 335.
The companies that poached its talent tended to be smaller AI labs or Meta SIL / Anthropic.
This is based on 24 months of LinkedIn-visible engineering moves in and out of OpenAI (Oct 2024 to Oct 2026).
OpenAI hires about 8 engineers for every 1 it loses. Against almost every company on the chart, it wins.
4 patterns stood out:
1. Big Tech is the farm system.
- Meta sent 312 engineers, more than anyone.
- Apple sent 211 and took back 2.
- Google sent 196.
2. Product companies feed it too.
Stripe sent 65, Netflix 50, Airbnb 43, Snap 41, and Robinhood and Uber 34 each. I ran Stripe's numbers on Monday, and they match: 64 engineers left Stripe for OpenAI from Stripe's side, 65 arrived from OpenAI's.
3. The losses go to smaller labs.
Anthropic hired 31 OpenAI engineers, and OpenAI hired 11 back. Thinking Machines Lab took 13 and sent 1. Reflection AI (5 to 1), Applied Compute (3 to 0) and Periodic Labs (2 to 0) round out the list. Those five are the only companies I checked where OpenAI loses more engineers than it gains.
4. Meta is the biggest two-way trade.
Meta sent 312 and took 38, which still makes Meta OpenAI's top destination.
OpenAI doesn't have a problem with its engineers leaving for Big Tech. Its leavers tend to pick smaller labs where each engineer owns a bigger piece of the pie.
I'm watching this closely to see how things change over the next 24 months. Follow for more breakdowns.
11
462
Cognition and Factory spent September fighting on X after Factory CEO Matan Grinberg accused Cognition of poaching one of its board advisors, the former CRO of Snowflake.
The hiring data says it's not a fair fight.
I was curious so I pulled every engineer at both companies. SWE, research, MTS, eng leads.
- Cognition: 32 a year ago. 78 now. +144%.
- Factory: 20 a year ago. 41 now. +105%.
Over the last 6 months alone, Cognition added 29 engineers. Factory added 12.
"But what about the Windsurf deal last year?"
Only 6 of Cognition's current engineers came from Windsurf or Codeium.
Cognition also has 77 people with "deployed" in their title. That's about the size of its entire core engineering team. Factory has 32 in solutions and deployed roles.
Where they recruit:
- Cognition: Google, Scale AI and Nuro (9 each), Jane Street (7)
- Factory: Google (6), Meta (5), Microsoft and Scale AI (4 each)
On Sep 30, Matan Grinberg, who co-founded Factory with Eno Reyes, posted that Factory was "terminating Chris Degnan for unethical conduct involving Cognition."
Scott Wu replied less than two hours later: "competing with integrity doesn't include posts like yours." Degnan says he resigned and told Matan he was going to Cognition.
Matan also wrote that Factory's team "has 10x'd in size." On LinkedIn, Factory's engineering team doubled.
Coding agents aren't a demo race anymore. The winner is whoever gets their forward deployed engineers inside the most customers, fastest. Right now that appears to be Cognition.
Who do you think wins this one?
4
49
10,190
Stripe poaches its engineering hires mainly from Amazon.
Who does it lose to? Frontier labs like OpenAI and Anthropic.
I looked at the data and 262 engineers joined Stripe from Amazon. 107 left for OAI and Anthropic.
This is based on 24 months of LinkedIn-visible engineering moves in and out of Stripe (Oct 2024 to Oct 2026).
As a former ML engineer in Big Tech, I've always been fascinated at studying where top talent flows, and now that we're hiring like crazy at Hellyeah AI, we often interview candidates from these companies.
But back to Stripe. Our analysis showed 4 interesting patterns:
1. Amazon is Stripe's biggest feeder by a mile.
Amazon and AWS sent 262 engineers, 18% of every outside engineering hire. Amazon took back 5. Microsoft came second with 75.
2. Fintech feeds Stripe too.
Block sent 22, Coinbase 16, Capital One 15 and Shopify 11. Stripe sent 8 back to those four combined.
3. Frontier labs are the top choice exits.
OpenAI hired 64 Stripe engineers, more than any other company. Anthropic hired 43. Those two took 14% of every Stripe engineer who left for a new job. Across OpenAI, Anthropic, xAI, Google DeepMind, Perplexity and Thinking Machines, Stripe hired 4 engineers and lost 118.
4. Google is the only even trade.
Stripe hired 27 Google engineers and lost 22 to Google. Meta ran 35 in and 16 out.
My read here is that stripe trains the kind of engineers AI labs want - people who ship systems that can't go down. Stripe out-hires Big Tech, but it can't out-pitch the frontier.
Which company should I map next?
3
2
585















