Rahul Kayala

@praneetdutta @GoogleDeepMind @HelloPomo Congrats guys!

I quit @GoogleDeepMind to build a marketing team that never stops working. Today, we're launching @HelloPomo. Your marketing, from insight to execution, unified. Pomo watches everything happening across your market, tells you exactly what to do next (the budget shift to make, the trend to jump on, the buyer to answer), then prepares the assets for your approval. @JoeC7007 and I have raised a $4.5M Seed led by @KindredVentures (@stevejang ), with backing from @SevenStars_VC (@stevenl ), @svangel (@AndreaShuyuWang ), @databricks, Timeless (@gauravahuja, @eshan_shetty), @645Ventures, @scottbelsky, @mghissassi & @mascarom. Reply with your website. We'll run Pomo for 10 companies and show you the growth opportunities it finds and the work it creates.

@JayaGup10 Great take. I think the runaway spend and corresponding lack of results (which you frame brilliantly as too much intelligence) is their own doing.

Hot take: OpenAI may actually have a shot at enterprise vs Anthropic. Pattern across some F500s right now: ChatGPT as the org-wide default, Claude ring-fenced for power users because of 1.) variable-cost fear and 2.) “more model than the median employee needs.” The first is a trap of Anthropic’s own success. Claude’s identity is welded to agentic workloads like Claude Code, autonomous multi-step work. The horror-story AI invoices circulating in CFO Slacks are tied to Anthropic (for now). The second is more damaging. “Too smart for the median employee” means frontier capability stops being the purchase criterion for 90% of seats, and a capability lead stops converting into distribution. The second-order effect: the default surface accumulates the org’s connectors, permissions, and working context and maybe advanced users eventually converge on wherever their team already operates. A possible end state: “our OpenAI relationship,” is board-level, vs “our Claude spend,” reviewable and cuttable. What happens if you don’t have to win the model race to win enterprise, you just have to win the default?

@nikunj @tbpn 100%

@tbpn Petition to officially rename Ramp as “TBPN sponsor Ramp”

One take on OpenAI's leaked top 30 customers who have used over 1 trillion tokens: It's 30 companies in wildly different markets. If they were all LLM wrappers, that would be worrying. But instead it's companies like Indeed, Duolingo, TBPN sponsor Ramp, Shopify. These companies are touching very different pieces of the market. They aren't competitors. Might be one indicator that the AI economy is less circular than how the bears are framing it.

Echelon emerges from stealth with $4.75M to replace enterprise consulting giants like Accenture and Deloitte with AI agents trained on real expert knowledge. 🤖📊 Full story: venturebeat.com/ai/echelons-ai… #ArtificialIntelligence #EnterpriseSoftware #FutureOfWork @_EchelonAI

@sumukx solving their own problems first - great observation!

i know openai employees are truly using codex internally because when i give it a set of gpus and ask it to try an experiment, it knows to launch on a subset and validate before launching an entire job these little "wisdom moments" make CLI agents go from cool toy -> woah

This, not search, is the real Google killer. Truly helpful sidekick that pushes the right information at the right moment!

Not @Benioff with the YC founder always be launching vibe!!

@GergelyOrosz I’ve seen this work well mostly for new grads and early career folks. Quite hard to get senior engineers to interview, let alone spend a week away from their day job

The sooner you realize interviewing is a series of tradeoffs, the better you will be. Yes, 60-minute algorithmical coding interviews do not represent the job... but they are very time efficient A 1-week trial perfectly simulates the job... and it's not time efficient!

I keep saying that (paid) trial weeks are becoming the only type of interview where you can do both things: 1. Simulate the real work, no algo interviews 2. Do not care one bit if candidates use AI coding tools: in fact, encourage it! Linear been doing this for years...

They can absolutely be founding team members. Just be honest about the trade-off. Give them equity, pay them well, but don't expect them to stick around during the boring months when there's nothing sexy to post about.

Your "founding creator" will quit the moment they hit 100K followers. They're building THEIR brand, not yours. The content that goes viral about your startup? It's actually about them.

Hot take: The "founding creator" trend is a trap most startups are walking into blindly. Every startup now ships cinematic launch videos because when software takes weeks to build, distribution buys you some moat. But we're confusing skills with DNA. 🧵

@JayaGup10 Because FT as a service would sound like saying fatass

why did finetuning as service never take off

@rak_garg MIT posted a staggering stat and here’s how my AI company is in the 5%

mom will you come get me LinkedIn is still talking about that 95% of AI fails study from MIT

Conferences can be pretty effective channels to land and close customers. Avoid at your own risk!

@lulumeservey @astronomerio well I hadn’t heard of astronomer before today so it’s not all bad news

Thoughts for the @astronomerio team Don’t bother with crisis comms here. The CEO will try to get you to protect him, but it’s not his company. He’s a temporary steward. Your job is to protect the company The CEO is a professional manager who’s only been there two years, the HR person has been there less than a year, neither is tied to the identity of the company Preserving trust is more important. Your business model is to handle sensitive data and your priority is scaling, and that’s tough to do with multiple known liars on the senior leadership team A better comms plan than trying to save this situation: Andy Byron is on the board, but he’s not a founder and doesn’t have control. The other five board members should replace him You can then use the new CEO announcement as a reset, and get people to focus again on Astronomer’s actual business instead of its drama

@chamath @dwarkesh_sp dis you!!

Can someone please start CAA for AI people.

@nikunj You need to do the work to get a reward and improve your decision making

Given the speed at which we're becoming bionic, this is a question we'll be asking ourselves every day..

@harris would be great to read once done

I'm rewriting the Series A Guide. Markets have changed in the last five years and advice should change along with it. If you've raised an A in the last 6 months, I'd be grateful for 30 minutes of your time. DMs are open.

@khoomeik This is beautiful dude!

devanāgari tattoos are cringe and played out but there’s a lot of alpha in siddhaṃ calligraphy born in the Gupta Empire, transmitted through the Tang Dynasty, and perfected in aesthetic form by the Japanese Buddhists