partner builder @645ventures | investing in AI agents, dev tools, and infrastructure | previously @initialized @fcollective @mit | here to help founders win
Joined April 2007
- Tweets 4,643
- Following 3,515
- Followers 11,591
- Likes 46,355
299 Photos and videos
parul retweeted
Jun 8
Any developer with fewer than 2M app installs can now get frontier-level model performance in their app. For free. Integrated with the existing FoundationModels API.
That ought to be the headline out of WWDC.
24
55
1,400
130,763
parul retweeted
Jun 8
It's not FAANG anymore.
It's MANGO.
642
1,629
25,913
4,150,398
parul retweeted
Jun 8
one of the joys of my life is placing talented folks in the right roles on the right teams.
for this edition of early days, we're highlighting some favorites on the market right now.
let us know if you want to hire any these fine folks!
8
6
134
17,201
đź‘€
in today's keynote, apple produced this really interesting graphic that ironically outlines the core mechanics for a new type of operating system (for perhaps a new class of devices).
you can see how this moves the world from an app based ecosystem to an intent centric world.
i.e. you roughly do not need third party applications in this world at all esp when ai has the ability to construct & deconstruct interfaces / experiences on demand.
1
2
200
How to AI by the queen đź‘‘
Jun 8
ok so i've spent years thinking about what it takes for execs to AI-pill their teams. a few thoughts on where i think leaders should start:
đź’» Technical Readiness
1. Stop chasing incremental velocity; you can 2-3x PRs if you try (h/t @darraghcurran)
2. Make it easy for jr engineers to onboard. your agents will thank you.
3. Increase investment in platform teams (devx ftw)
4. Focus on verification loops, not prompt engineering - we love a /goal
5. Speed up core feedback loops (lints, tests, etc.)
6. Figure out how to bypass human code review for some PRs
7. Do extreme eng experiments (delete your IDE! h/t @chintanturakhia)
8. Ensure your team can experiment w models, harnesses, etc.
🤝 Operating Model
9. Put AI workflows into buckets: automate / augment / ignore / kill
10. Turn your best AI user's flows into a system- i think @JJEnglert is good at this
11. Give agents actual jobs, not tasks
12. Get non-engineers committing code to production - use @DevinAI if you need a place to start
13. Measure tokens (it works, sorry not sorry)
14. But have a controlling quality measure
15. Question every process built to protect scarce engineering time
16. No pure managers
17. Create new team topologies (v small! very large! v flat!) @GammaApp and @thisisgrantlee @thatsjonsense do this well
🥳 Culture
18. Require leaders to build something with AI and demo it publicly
19. empathaize: your team fears looking incompetent without AI AND replaceable with it
20. Measure AI fluency quantitatively (I have a survey for this)
21. Name AI champions and reward them like @doshkim
22. Learn how to hire for AI proficiency
23. Don't be a CEO that delegates understanding AI to someone else
24. Create informal and formal reward systems
25. Have fun (my mantra)
Sound good but need help? join 100s of CTOs, VPs of Eng, Product, and Design learning how to do this for real during our 2 day @MavenHQ workshop.
25% off till midnight tonight! → maven.com/clairevo/ai-native…
393
parul retweeted
Jun 7
the anthropic co-founder jack clark advice that stuck with me:
read the primary material. not the summary. not what the ai said about it. the actual thing.
form your own opinion first. then ask the model. never the other way around.
keep practices in your life where it’s just you against the world ~ a sport, an instrument, reading, building something with your hands. spaces where the algorithm can’t mediate what you learn about yourself.
and don’t defer to AI even when it’s usually right. especially then, actually. that’s precisely when the habit forms.
the people who won’t get eaten by this moment are the ones who stayed hard to replace. not because they avoided the tools but because they kept the parts of thinking that make the tools worth using.
53
256
2,331
123,002
parul retweeted
Jun 6
There are only 4 PR strategies:
1. The Hunter Biden Rebrand:
- Laugh at your mistakes
- Lean into the drama
- Become a character instead of a scandal
2. The Corporate Apology:
- Hire consultants
- Post the notes app
- Apologize profusely
- Kiss the ring and hope forgiveness comes.
3. The Churchill Justification:
- Own the decision, reject the shame.
- Explain why it was necessary and let history decide if you were right.
4. The Trump Counterattack:
- Make the story about your enemies
- Attack so aggressively nobody remembers the original accusation.
19
95
1,403
96,644
parul retweeted
Jun 4
they fled at exactly the speed limit
19
26
527
37,045
parul retweeted
The thing nobody tells you about exponential change is that it feels like nothing is happening right up until the moment everything happens at once.
677
2,190
12,169
16,307,757
parul retweeted
Jun 3
When the cost of doing something at scale goes to zero, the value of that thing goes to zero too.
AI SDRs will stop working the moment everyone has one. Now the human salesperson has become the premium signal again. Enterprise deals are being closed on calls and text.
The things that don't scale are going to become more valuable as everything else does. Actual presence, carefully nurtured relationships, real conversations, etc. The pendulum always swings back.
20
6
80
7,792
parul retweeted
Jun 2
SF was over if not for OpenAI and Anthropic!
Before that, it was over if not for Stripe, Uber, Airbnb, Twitter, Salesforce, Visa, Gap... and if you go back a bir further - Levi's, Bank of California, Spreckels Sugar, Wells Fargo, Union Iron Works, Southern Pacific :)
The city gets “saved” by the next boom every time. After the gold rush, there was silver, railroads, shipping, banks, sugar, utilities, Pacific trade, defense, semis, PCs, biotech, enterprise software, the internet, social, mobile, SaaS, fintech, crypto, and AI.
I think that's just how it goes in this town.
Jun 2
It probably was over if it wasn’t for OpenAI / anthropic.
59
58
1,289
163,426
Want to try GStack /office-hours for your product idea as quickly as possible? It's one click here:
app.lightsprint.ai/office-ho…
64
39
662
75,180
parul retweeted
Jun 1
sf has tons of free events but they're buried across funcheap, eventbrite, reddit, random instagrams. i'm always looking for something fun to do in sf and i'm sure you are too!
built a map that pulls them all into one place - comedy, workshops, art openings, free food, pop-ups. updates daily... check it out! 🧡
sfrats.com
90
83
1,752
122,727
Creative but unsurprising attack vector
Jun 1
It’s wild how Meta - a company going all-in on AI - somehow missed the memo on how AI can generate images and videos that renders “take a selfie of yourself” verifications utterly useless
So now Instagram accounts hacked at scale. 2FA also fully bypassed - by Meta’s own design
1
381
Anthropic headed to IPO with craziest revenue ramp in history:
2021: Founded
2022: $10 million
2023: $100 million
2024: $1 Billion
2025: $9 Billion run-rate
January 2026: $13 Billion run-rate
February 2026: $19 Billion run-rate
March 2026: $30 Billion run-rate
April 2026: $44 Billion run-rate
May 2026: $47 Billion run-rate
Jun 1
Anthropic has confidentially submitted a draft S-1 registration statement to the Securities and Exchange Commission.
Pending completion of SEC review, this gives us the option to pursue an initial public offering.
Read more: anthropic.com/news/confident…
21
53
712
131,396
parul retweeted
Jun 1
I don’t think any of you understand what is about to happen in the market. We are about to live through the craziest five year run in technocapital history. God help us all. I pray that when Judgement comes He can see all that we did to ensure efficient price discovery.
44
83
2,199
240,454
parul retweeted
May 28
someone hit me up about the new "claude dynamic workflows" feature, claiming "see, multi-agent works"
But really, the launch of this feature proves the exact point that I made back in June of 2025, along with @walden_yan, @tobi, @karpathy, and many others:
Deterministic workflows orchestrating small agent loops beats non-deterministic multi-agent or "agent soup" systems every dang time
everything is context engineering
Excited to share our most powerful new Claude Code feature: dynamic workflows!
Mention "workflow" in a prompt and Claude will dynamically create an orchestration plan that it strictly follows, allowing you to confidently trust that every stage happens in the right order even across 100s of agents.
75
80
1,092
203,609
If you work in the software industry and have time to read only one long-form post today, read this one.
If you have time to read two, read this one twice.
Highly #recommend
tl;dr: Stay off the yellow brick road that the frontier model companies are racing down. There is plenty of opportunity to solve hard problems elsewhere. Focus on areas where you can build the system of work (workflows), capture compounding, non-public data and deliver deterministic outcomes that customers need.
28
154
1,768
460,384