Y Combinator

Paul Graham (@paulg) whether founders should move to Silicon Valley, and what it takes to build a startup hub anywhere else. Live from our YC | Stockholm event on April 29, 2026. 01:01 – Why the Big Center Matters 02:45 – The Power of Serendipitous Meetings 04:36 – Investors Move Faster in the Valley 06:03 – Respect Follows the Move 07:59 – The Dropbox Story 09:10 – Measuring Yourself Against Big Fish 12:21 – Silicon Valley's Pay-It-Forward Culture 15:36 – How to Help Stockholm Thrive 17:24 – YC as the Optimal Path 19:54 – Could Stockholm Become The Silicon Valley of Europe?

Was fun to host @jerallaire, founder of USDC stablecoin-issuer @circle, for a fireside chat @ycombinator in San Francisco. We touched on everything from agentic commerce, to better worldwide stablecoin regulation, to his founder journey in a regulated space. Some highlights 👇🏽

The $2T telecom industry was built for humans. @AgentPhoneHQ is rebuilding it for agents. One API gives every AI agent its own phone number and a trusted identity to reach the real world. Congrats on the launch, @themeetmodi & @manav2modi! ycombinator.com/launches/QNE-a…

17 years ago I sat in a Palo Alto garage and interviewed for YC funding — during the depths of the Great Recession. Recently, I went back to that house and talked to @paulg and @jesslivingston of @ycombinator about building things that survive. First episode of the @disasterproofHQ podcast --> youtu.be/KBjtRJSblMU

I sent this message to our management team last week: Pylon managers must become AI-native. What becoming "AI-native" means for managers, is being able to increase 𝗾𝘂𝗮𝗹𝗶𝘁𝘆 of your work and 𝗾𝘂𝗮𝗻𝘁𝗶𝘁𝘆 𝗼𝗳 𝗱𝗶𝗿𝗲𝗰𝘁 𝗿𝗲𝗽𝗼𝗿𝘁𝘀 you can effectively manage. We think this is especially possible in customer-facing roles today (support, success, sales, solutions). So how does one become an "AI-native manager": 𝗦𝘁𝗲𝗽 𝟬: 𝗚𝗲𝘁 𝗼𝗻 𝗖𝗹𝗮𝘂𝗱𝗲 𝗖𝗼𝘄𝗼𝗿𝗸 𝗼𝗿 𝗖𝗵𝗮𝘁𝗚𝗣𝗧 𝗦𝘁𝗲𝗽 𝟭: 𝗖𝗼𝗻𝗻𝗲𝗰𝘁 𝗲𝘃𝗲𝗿𝘆𝘁𝗵𝗶𝗻𝗴 𝘃𝗶𝗮 𝗠𝗖𝗣 If you haven't starting connecting your systems of record via MCP, do so immediately. This will allow you to build agentic workflows and dashboards. At Pylon we've connected 50+ products at this point from Salesforce to Pylon to Gmail... 𝗦𝘁𝗲𝗽 𝟮: 𝗥𝗲𝘁𝗿𝗮𝗶𝗻 𝘆𝗼𝘂𝗿 𝗯𝗿𝗮𝗶𝗻. Don't use LLMs as a replacement for Google. Instead, think about if you had infinite time, how would you become the best manager? For example, a sales manager might... → Review every sales call and prospect interaction → Give feedback to AEs when they miss talking points → Knowing which AEs struggle against which competitors → Pro-actively help AEs re-engage CL deals → Review all categorizations of every opportunity (e.g. "why they booked") Turns out... this is all possible now. At scale. I prototyped something like this for our AE team to prove the point. I asked Claude to... → Go through every single closed deal, and label why they booked, where they heard about us, what resonated, ... and to write that data back into Pylon + Salesforce → Listen to every sales call, and find which questions our team isn't great at answering → Create dashboards to visualize this data across the entire team → Find closed lost opportunities that we should re-engage based on deals lost due to missing functionality 𝗦𝘁𝗲𝗽 𝟯: 𝗗𝗼 𝟭𝟬 𝘁𝗵𝗶𝗻𝗴𝘀 𝗮𝘁 𝗼𝗻𝗰𝗲 Because you have access to agents, you can queue up many asks at once that will take a while to run. Customer support teams are used to doing this with multiple open tickets at a time. Engineering teams started doing this with Cursor/Claude Code. When delegating long-running tasks, you have to get good at parallelization. 𝗦𝘁𝗲𝗽 𝟰: 𝗖𝗼𝗺𝗺𝘂𝗻𝗶𝗰𝗮𝘁𝗲 𝗶𝗻 𝗺𝗮𝘀𝘀 𝘁𝗼 𝘆𝗼𝘂𝗿 𝘁𝗲𝗮𝗺 One of the most underrated ways to use AI when managing many people is to start using it to send messages to your team via Slack (via Claude/ChatGPT). This is way more powerful than I thought. For example... I wanted us to find lost opportunities to re-engage. So I first asked Claude to check Notion + Linear for new product releases. I then asked it to comb through lost deals and check account context in Pylon to see if we lost the deal due to any functionality we now released. I then asked Claude to DM 13 AEs a list of their lost deals to re-engage with context on why we lost them. It's magical. [I've hit the LinkedIn character limit, more soon]

The biggest alpha leak of 2026 is that you can tokenmax $10k/mo with OpenClaw/Hermes + GBrain and get the AI that everyone will have in 2028 for $100/mo, but you can get it now, and that is the biggest single unlock you can have vs your competition

We've partnered with the state of Texas to launch efficiency.texas.gov, the nation’s first AI government concierge for laws, regulations, licenses, and permits. SAM AI lets citizens and businesses search across the Texas legal code, register and renew licenses, and fill out permits, all in one place. We are proud help Texas make its laws and permits accessible to all. gov.texas.gov/news/post/texa…

Back at it. At the mothership. Grateful to be amongst the world's best and brightest founders at YC. @garrytan as group partner. He was also Webflow's group partner when we both had less white hair. Can't wait to show ⬜⬜⬜⬜ to the world

We're announcing our $2.7m seed round from @bwcvc , @ycombinator, @tinyvc , and @venturetogether to build the autonomous tax firm 🌵 Walked around the nearest park to explain why companies like Hadrian, @openfx_ , @verticeHQ , @ButternutBox , @encord_team , @DuffelHQ , and @Immutable pick us over law firms & tax advisors for their Transfer Pricing. We're growing fast. Surpassed '25 ARR in Q1 alone so we're hiring across tax + product teams - if you're a transfer pricing nerd, please get in touch! Thanks to my amazing team, friends, investors, and angels for believing in us 🤙🏼

Yeesh.. is there no benchmark @poetiq_ai can't dominate?

From factory floor to installed on stand in less than 18 hours - not bad for SN1!

the 🐐@snowmaker

Tune in: youtu.be/ZVdgiuiOQ7E

Jeremy Allaire (@jerallaire) is the Chairman, CEO, and co-founder of @circle, the issuer of USDC — a stablecoin with nearly $80 billion in circulation and a core piece of the broader stablecoin market, which has grown from zero to more than $300 billion over the past decade. In this fireside, YC Visiting Partner @nemild sat down with Jeremy to talk about Circle's founding vision of building an internet protocol for dollars, the wave of institutional and regulatory adoption now reshaping the financial system, and why agentic economic activity, not just consumer payments, might represent the most transformative frontier for stablecoin builders in the years ahead. 00:00 - Why Stablecoins Matter 01:45 - Circle's Origin 07:10 - Top Use Cases for Builders Today 11:22 - Consumer vs. Business Adoption 13:20 - Banks & Institutions Enter the Space 17:54 - Global Regulation Landscape 22:40 - AI Agents & the Agentic Economy 26:15 - Missing Infrastructure for Builders 27:38 - Three Big Predictions for Stablecoins

Astraea brings drugs to market faster by fully automating clinical trial biometrics, turning raw study data into CDISC-compliant datasets, TFLs, and FDA-ready outputs in days. Congrats on the launch, @joshwqngsr and Sanmay! ycombinator.com/launches/QLz-a…

we built the first sane way to debug your agent locally. you can see your traces. codex/claude code can too. this lets them write evals and test your agents automatically. best part: it's completely free and open source. install with 1 line. (github below)

PerfectBit (@PerfectbitAI) creates high-quality training data that's correct by construction. They verify against physics simulators, scientific databases, formal proof systems. LLMs, robotics, AI for Science, and more. Congrats on the launch, @vaphab and @sjy82155065! ycombinator.com/launches/QMv-p…

@AseonLabs is honoured to be the first feature in the Ground Truth series from RideAI. A great new series focused on the infrastructure layer of autonomy, which happens to be exactly what we are working on. Big shout to @mikemallazzo for a quality interview. rideai.org/introducing-as…

Clara (@askclarahealth) is an AI-powered primary care doctor that reads your whole medical history, diagnoses, and treats - with licensed clinicians reviewing every decision. From the team that built Circle Medical to nearly 1 million visits/year. Congrats on the launch, @georgefavvas, Zeeshan, and Caitlin! ycombinator.com/launches/QMs-c…

.@TakeCareOS is an AI-first operating system for care agencies. It consolidates every function an agency runs on — rostering, shifts, CRM, care notes, messaging, invoicing, timesheets, and compliance — into one place, with AI agents that automate the admin work coordinators currently do by hand. Congrats on the launch, @ragavsachdeva! ycombinator.com/launches/QLS-t…

Tune in: youtu.be/RMmEB9ajmdo

We asked a dozen DevTool founders from companies like @RevenueCat, @greptile, @firecrawl, @infisical, @ollama, @resend, @mintlify, @UnslothAI, @porterdotrun, and @recallai, about the state of AI agents and the future of software engineering. In this episode of Founder FAQ, we covered everything from agents as customers and the end of coding, to advice for founders starting out and what they're most excited about going forward. Their answers might surprise you. 00:00 – Meet the Founders 03:00 – Building for Agents First 04:22 – Biggest Early Mistakes 07:15 – Do Founders Still Write Code? 09:22 – Most Unexpected AI Discoveries 12:09 – What's Underrated Right Now 14:38 – Predictions & What's Next