🦄 Why payroll is still broken

👋 Howdy to the 1,026 new legends who joined since our last edition! You are now part of a 578,878-strong tribe outperforming the competition together.

LATEST POSTS 📚

If you’re new, not yet a subscriber, or just plain missed it, here are some of our recent editions.

📔 The Fine Art Of Disagreeing & Committing. Why consensus is where good decisions go to die, and the seven-step fix.
🌏 How Product Teams Are Going Global In Days. An interview with Minh Cung, Co-Founder & CEO at Hyperlocalise.
📞 Why The Next Interface Can Hear & See You. An interview with Alex Smola, Co-Founder & CEO at Boson AI.

PARTNERS 💫

Everyone’s talking about getting back to the office. Candidates are still searching for remote.

Ashby’s data from 1,200+ startups show that remote roles draw 42% more applications and 9% higher offer acceptance rates, even as remote jobs slipped from ~80% to ~60%. For more insights, read our free State of Startup Hiring report.

Business banking was built for people to manage. Airwallex's agentic business accounts manage themselves.

Powered by global financial infrastructure, your AI agents can now provision local currency accounts, execute cross-border payouts, and optimize liquidity across entities 24/7. Put idle cash to work and move funds at the speed of your growth, all within the rules and approvals you define.

_{Interested in sponsoring these emails? See our partnership options here.}

HOUSEKEEPING 📨

The final Dune: Part Three trailer dropped this week. It’ll be super interesting to see what Denis Villeneuve does with this final installment. It’s no secret that Paul is kinda building towards being the tyrannical bad guy (I’ve actually read the books, so won’t spoil).

One thing I’ll never understand, though, is how Chalamet, who I usually love, got cast as Paul. No offense to Timothée, but if it were Lord of the Rings, he’d play Frodo, not Aragorn. I find him so hard to believe as the leader of an interstellar holy war. Anyway, pumped for the final movie, nonetheless. I think it’s not LOTR, but damn it’s a good franchise.

INTERVIEW 🎙️

Ayush Sharma, Founder & CEO at Warp

Ayush Sharma is the Founder & CEO of Warp, an AI-native employee management platform that handles payroll, tax compliance, benefits, and IT for high-growth companies. He studied computer science and physics at MIT, stayed on for a master's focused on systems and machine learning, then worked as a software engineer on Yelp's core machine learning team. In 2023, he launched Warp out of Y Combinator's Winter batch. The New York-based company now serves more than 1,000 customers and is on track to process over $2B in payroll this year. In June 2026, Warp announced a $60M Series B led by Battery Ventures, with Shopify CEO Tobi Lütke and former Stripe COO Claire Hughes Johnson joining the round, less than a year after an $18M Series A led by Sound Ventures. Total funding now sits at $85M.

Ayush came up with the idea for Warp at a fast-growing startup where he ran payroll himself and lost hours to state tax filings and other rote work that software couldn't automate. His bet is that the U.S. tax system, split across more than 10,000 jurisdictions and still running largely on paper and phone calls, is exactly the kind of mess AI agents can finally clean up. Early on, investors assumed payroll was a solved category the incumbents already owned. Today, Warp serves its customers with the equivalent of 1.5 people on its tax team, and it put together its preemptive Series B in six days.

Why are payroll and compliance still broken for startups?

I think this question cuts right to the heart of the problem. People on the outside often assume that because every business has to run payroll, it must be solved and working by now. Then you get on the inside and find out how much of it is duct-taped processes and manual work. Much of the software supposedly automating the work is really just a fancy dashboard that surfaces the task back to a human, who then has to do it and take liability for any errors. That's maybe the non-obvious point. As for why it's still broken, I'll give you one concrete reason that's specific to the United States. At Warp, we primarily serve U.S.-first companies that are expanding across the country.

Part of this is historical. It goes back to how the United States was structured, with a real focus on decentralizing power so everything didn't get concentrated at the federal level. I grew up in India, where the central government controls most taxes, and everything is set in one place. In the United States, there are over 10,000 tax jurisdictions. It's really crazy. And if you've ever been to a .gov website, these are very old tax departments that mostly run on phone calls, paper, and notices that arrive in your mail asking you to do certain things. Their websites all suck, unfortunately. As you can imagine, getting anything done there is really complicated.

So the short version is that it's broken because there's so much archaic fragmentation in the US tax system. What we've set out to do at Warp is point a new technology, AI and LLMs, at this very messy, complex, non-software category of problems. We think it can finally automate what the last generation of traditional software couldn't. That's why it's still broken, and I think we can fix it.

What compliance work does Warp automate that competitors tend to ignore?

Some quick background first, because two parts of my story matter here. First, I studied computer science, math, and physics at MIT and ended up specializing in and researching large-scale machine learning systems. After graduating, I worked as a software engineer and machine learning engineer for a while, continuing that research work. The second is that I joined a startup because I wanted to build more and take on more responsibility, which usually isn't possible at a large company. It was a Series A company, and I joined when they were around 10 or 12 people. In less than a year, they grew to 40 or 50 people, about 4x or 5x in a short span. I personally hired and onboarded a bunch of people, and I realized how much manual work was still hiding behind the first layer of software in this space.

To give you a concrete example, take New York, where I'm based. It's actually one of the more frustrating places to deal with for taxes because it has four or five different tax jurisdictions. There's the Department of Labor, Withholding, Paid Family Leave, Disability Insurance, and so on. Most people miss these until a tax notice shows up in the mail. 'You owe us $10,000 because you missed this one random filing, which you were supposed to do on our website’—which doesn't work, by the way. People on the outside assume this would be solved by now. 'How the hell have you not figured this out? It's 2026.' But it turns out that's the thing no other vendor or competitor has been able to solve. The fundamental reason is that everyone has built with traditional software, which works for deterministic systems, nice APIs, and software-type problems. When you're dealing with government agencies, tax notices, and tax regulations, traditional software just can't solve it.

So when you hire someone on Warp, unlike with any of our competitors, you never have to think about or touch any of the .gov tax compliance notices. If you're a high-growth company hiring a bunch of people and you're using anyone else, that's basically half your job. You keep up with tax notices, file on time, and make sure you're registered in the right places so everything goes smoothly. Otherwise, you face errors and bills that take a long time to unwind. It's frankly such a drain on your company's operations. That's the kind of work we do. It was one of the first big areas, and from there we're obviously building out an entire platform now, but it's the core thing we try to do way better than everybody else.

Why build an employee management platform that's AI-native from day one?

We realized that traditional software systems in this space, like traditional HRIS platforms, assume software ultimately exists to help a human track a task or do it a bit more efficiently. Our bet at Warp is that this model won't survive the next couple of years of AI. You're starting to see that in a lot of areas. Software engineering, interestingly enough, is one of the clearest examples, which is funny because people thought it would be the last thing to get solved.

It turns out it's one of the first, because certain characteristics of the problem lend themselves really well to AI. Internally, our model with our engineering team has completely changed. Our job is no longer just to hire software engineers. It's become much more about architecting our product, engineering, and design teams so AI can do a lot of the work internally. Agents are building entire features for us, and we're the commanders of the ship, deciding what to build, how to test it, and more.

Tying that back to the platform bet, the product is the same idea. Our incumbents are trying to get humans to use their software more. Our goal is almost the opposite: how can we get you to spend less time in Warp? We automate the work in the background, set things up, and use AI as much as possible so your company has that power. You stay in control, review and validate everything, but we automate the tedious work you do day in and day out. That shows up in the product architecture itself. I think it's one of the reasons a lot of the larger companies now rushing into AI marketing, often by launching some random chatbot on top of whatever they already have, are going to have a very hard time. You can't just bolt this on as a thin chatbot. It has to be built deeply into the platform from day one.

What was the hardest part of Warp's earliest days?

We started a little over two and a half years ago. We got into YC, and I remember our YC interview was a super-fast 10 minutes where they tore us apart. YC has funded some successful companies in this space over the last decade or two, and those companies have made YC a lot of money. So that group of people is very aware of where things have and haven't worked in the past. I faced what I'd call a very brutal interview. 'Why do you think you can do this? What’s your background? What do you know that all these other people don't?' In some ways, those are actually the right questions to ask yourself when starting a startup.

Early on, many investors were skeptical, assuming the category was solved and incumbents would take it over entirely. For us, the hardest part of the early days was holding firm on our own conviction. We believed AI would allow for a fundamentally different kind of product here, way better and way more differentiated than what existed, and that the incumbents hadn't caught wind of what was coming. And we were the right engineering team to build that future. We did get customer traction pretty early. For as long as we can remember, we've never had a phase where we weren't drowning in product requests from our own customers. That was true when we had 10 customers, and it's still true now that we have over 1,000. It's one of the more fortunate things about starting Warp.

Now we serve a ton of companies, and we've raised a lot of money. We just announced our $60M Series B, led by Battery, with some really cool folks like Tobi, Shopify's CEO, joining the round. The investor side has come around as we've kept shipping more product, showing that customers love Warp, and building out the team. But internally, that early skepticism was always the hard part. That said, in the grand scheme, there are worse problems. I think the worst problem would be not having strong customer momentum and feeling lost. Thankfully, we've always had the opposite problem, with customers constantly telling us what they need us to build, fix, and improve. That's its own kind of problem, but it's a way better one than the alternative. I've always said that internally.

PARTNERS 💫

You take the calls, Granola takes the notes. Granola transcribes your calls from your computer while you jot down a few bullets, then writes them up. At the end of the day, you have notes on all six, not two good ones and four you meant to finish.

Get 3 months off any paid plan with code OPENSOURCECEO.

_{Interested in sponsoring these emails? See our partnership options here.}

How did you go after your first customers?

When we got into the YC batch, one cool thing was that YC is a trusted network of people who like to build things and are very proactive about adopting new tech. They like new products. It's basically a community of builders and tech enthusiasts, so it's usually easy to find a couple of folks willing to try something when you tell them you're building something that does the job better than everything else out there.

I'd tell them we were building Warp, that it solved all these tax compliance problems their current provider probably didn't, and ask if they'd like to chat and give us some feedback. We were pretty direct in those first calls, too. We'd ask them to sign up, start paying for it, and start using it. That was the initial mix, with me reaching out manually.

I do think it's one of the best ways. Put a signal out about what you're building, and if there's interest, reach out to those users directly as the founder. Tell them you'll personally onboard them and make sure they have a great experience, and that they should let you know about any bugs or feedback. That got us our first customers. Since then, it's obviously been a journey and we've done a ton of stuff, but that was the start.

What does your actual week as CEO look like?

I don't think I have a typical day or week anymore. I basically gravitate toward whatever's broken or the bottleneck in the company and try to fix it. There are inflection points in a company where the CEO's role suddenly changes, and I think I've found myself in the middle of one of those. The first was when we went from about seven people working out of a WeWork to a team of more than 10. When you're seven or eight people, you basically know in your brain who's doing what at all times. It's that simple. Once you scale beyond that, you can't keep all that context in your head. For the first time, you have to think about systems and how to make sure you're not needed for every single decision. That was a big shift, and it happened early on, shortly after our seed round.

More recently, we went from less than 20 people to 50 in about five months. We signed up for an office space we'd hoped would last us longer, but we've already filled it out, and now we have teams and execs. The other big shift around then was a lesson every successful founder and CEO learns eventually. To scale the company and the team, you need leaders who can build systems and lead the team, because I can't personally lead every function. In the early days, that's exactly what you're doing as the founder or CEO, so your instinct is probably to keep doing it, but it becomes a bottleneck very quickly past 15 or 20 people.

So, back to the original question. First, I manage my direct reports, and I only have about five. They're all leaders in the company, each running a functional area. I meet with them every Monday, and we talk through the company's overall state week over week and the key metrics we're tracking. Setting direction on what's working and what isn't is one of my most important jobs. It's a very tactical thing, and you won't hear about it in a lot of these conversations, but I think defining the metrics is one of the more important jobs a CEO has, especially in the Series A through C or D phase. You go over them every single week with the leaders responsible for them and run a loop. Is the number going in the right direction? If so, let's keep doing more of what's working. If not, let's change strategy and try different things.

Second, there's a certain amount of board and investor management, so let's call it investor relations. We just came out of a fundraise, and it was atypical because it was a preemptive round. Someone came to us with an offer, and we didn't actually go out to raise. We weren't planning on it, so it happened very quickly. We got some great folks on the table, and the round turned out really well. But even in a perfect scenario like that, there's still a lot of work coordinating with investors and negotiating the deal terms, the round structure, and the board composition. That stuff isn't very glamorous to talk about in these kinds of conversations, but it's a real part of company building.

Third, find the fires and leaks, go toward them, and fix them. The tactical thing I've found helpful there is this. Startups are full of fires at any given time, and there will be days when you think everything's broken. So you have to think through what's actually worth your personal time. I've found that the best way to prioritize is to ask what the company's actual bottleneck is today. What are the one or two bottlenecks that, if we fixed them, would get us to the next milestone and the next stage of the company, whether that's traction, the platform, or the numbers we're trying to hit? That's very helpful because it lets you filter and focus on what really matters.

That's not to say you should ignore the other areas. You still deal with them, and hopefully by this point, around the B or so, you have other leaders you trust to run and manage those problems too. But as a founder and CEO, I think it's very important to always force yourself to ask, 'Am I working on the most critical, most important thing I could be?' That's the exercise I try to do, at least, and it usually takes me in very different directions and across different teams, depending on what's going on at a given time.

What's your North Star metric right now?

On the qualitative side, our mission is to build the most powerful and delightful employee management platform for ambitious companies. We took some time to think through each word and its placement before landing on it, but it reflects how we think about Warp. Put in those terms, our goal is to become the default employee management platform for high-growth, ambitious companies with anywhere from five to 5,000 employees over the next three years.

How do you use AI day to day inside Warp, beyond the product itself?

I’m thinking about how to architect our engineering org to think as much as possible outside the traditional model, where you just hire more and more engineers, product people, and designers. We've never really done that. We have a pretty lean, efficient engineering headcount relative to the product surface areas we build, manage, and improve. But it's even more top of mind now. For example, we spend much more time on key platform architectural decisions, and much more time thinking about how to test and validate things, especially because we deal with high-stakes flows around money, taxes, payments, debits, and compliance. We spend very little engineering time writing the code itself.

So the first big part and the last big part are where our engineering team spends more and more of its time. Even there, we're thinking, 'Okay, if the bottleneck is now reviewing and validating the code, should we be investing more in AI to do the testing and validation continuously in the background?' These are the kinds of decisions we're making and building in real time. As an engineering-minded founder and CEO, it's exciting to live through this firsthand and build in it. I truly think it's the golden age of building.

One company I really admire is Stripe. Stripe's mission is to increase the GDP of the internet. It's very rationalist and SF-coded, but the sentiment underneath it is actually very positive and noble, and it resonates with me. It's about enabling more builders and entrepreneurs out there to build more things. Building new things is how everything around us has ever happened. Stripe did that for incorporation, taxation, and, of course, accepting payments. Those things ultimately lower the barrier to creating and building new things. At Warp, we see the next opportunity here. Our goal is to empower the next generation of companies to take that way further and eliminate an entire class of HR, finance, and administrative drag that slows them down. We're very excited about that.

How do you get the best out of yourself?

First, I really do think a big part of the CEO's role is managing your own psychology. So step zero is understanding that it's a key part of your job description as a founder or CEO. If you don't, that's when burnout happens, and people crash out. Being very candid, there are plenty of successful founders who end up doing too many drugs. There are cases like that. It's really hard to do. When you're ultimately accountable and responsible for every single thing that happens in your company, because the buck always stops with you, there's a lot of pressure. So the first step, I've found, is making peace with the storm.

I often tell myself, and anyone who comes into the office, half-jokingly, that the two together capture the founder experience in a nutshell. You're in choppy waters most of the time, but in your best moments, you rise above the fog, see clearly, and cut through everything. It's about reminding yourself that storms will come, and your job is to stay calm through them and keep building. That's the more philosophical side.

On the tactical side, I personally find saunas really incredible. I got into them about a year and a half ago and try to go pretty regularly. There are some good spots in New York. I've found them really helpful, especially when you follow a sauna immediately with a cold shower or a cold plunge. It's very relaxing, calms you down, and takes your mind off things. Working out, taking care of yourself, and eating well all really help too. But I've found the sauna is a big hack.

Extra reading

• Operating At Warp Speed - July, 2026

• Exclusive: Warp adds programmable agents to its AI-native HR platform - September, 2026

• Warp 2.0: Making HR Human Again with the Warp Agent - Unknown

And that’s it! You can also follow Ayush on LinkedIn and X, and don’t forget to check out Warp to keep up with what they’re building.

BRAIN FOOD 🧠

TWEETS OF THE WEEK 🐣

TOOLS WE RECOMMEND 🛠️

Every week, we highlight tools we like and those we actually use inside our business and give them an honest review. Today, we are highlighting Framer*—the site builder trusted by startups to Fortune 500.

See the full set of tools we use inside Athyna & Open Source CEO here.

HOW I CAN HELP 🥳

P.S. Want to work together?

1. Hiring global talent: If you’re hiring tech, business, or ops talent and want to do it 80% less, check out my startup, Athyna. 🌏

2. See my tech stack: Find our suite of tools & resources for both this newsletter and Athyna here. 🧰

3. Reach an audience of tech leaders: Advertise with us if you want to get in front of founders, investors, and leaders in tech. 👀