Today, we’re launching @TownAI: the AI assistant that learns you. We’re coming out of beta with a $55M Series A led by @ARampell at @a16z, with participation from @KirstenGreen at @forerunnervc and continued support from @firstround, @altcap, and @conviction. Right now, getting real value from AI means prompting, configuring, building workflows, managing agents. We think that’s backwards. The future of AI is a companion that already knows you and how you work. Town connects across your inbox, calendar, Slack, docs, messages, and workflows to understand what you need, then starts doing the work with you. Drafting. Scheduling. Project tracking. Follow-ups. Context gathering. Multi-step tasks. And it only acts when you say so. All adapting to your voice, priorities, routines, and relationships over time. Your Townie is the AI assistant you actually need.

We stopped relying on people to remember to write down product feedback after customer calls. Internally, we built a Customer Call Summary skill in Town that runs after any meeting with a customer. It takes the notes and if there’s product feedback or some kind of insight in there, it puts it in a Slack channel we call customer-insights. Then we have another thing that summarizes the channel. What’s cool is it doesn’t matter who had the conversation. It could be a salesperson, it could be a PM, it could be me having some random conversation with someone. If there’s something useful about the product in there, it ends up in the same place automatically.

I will now be CC'ing Maxie on all key email threads going forward, and releasing approximately 100 metric tons of cognitive load daily as a result. This team 🚢 🔥

Combined @clay + @TownAI today after seeing an example from @jgreze on X and this makes LinkedIn look like a baby toy for researching people and companies.

Launching CC your Townie today, which lets other people work directly with your Townie over email. It's an incredible time saver and will wow you in so many ways. Try it out. (I'll be doing a lot more posting about how I use this to save me hours every week and why trusting your AI to work directly with others is one of the biggest shifts in HCI)

Apparently Town is a hotel. A couple weeks ago some customers complained about charges they didn’t recognize. Turns out their credit card statements were categorizing Town as a hotel. That's fine. We can fix the Stripe descriptor. The problem is I really want Town to show up on the Ramp list of fastest growing apps by spend, and now I’m wondering if Ramp thinks we’re a hotel too.

Muse is fantastic. Maybe here they are a bit too excited about launching integrations everyone else has had forever. At the same time I love Alex's X game. Respect.

We asked Claude to make changes to an email marketing campaign. It created a duplicate and double-emailed some folks. @TownAI would never!

A hotel in the middle of nowhere, no internet, but cable, watching SportsCenter, late into the night, then Baseball Tonight, a great evening, always.

We’re using Town for candidate sourcing a lot internally right now, and it’s really good. You give it the role and it asks you a bunch of questions about what you’re looking for. Then it builds a sourcing plan and goes to Clay to find the kinds of companies you want and people at those companies. Which might be a little weird because Clay isn’t a recruiting tool. But Clay has companies, people and titles, so it turns out to be pretty effective for this. Then it does more research on each person and asks other people’s Town assistants if anyone on the team is connected to them. If they are, it drafts the email. If they aren’t, it suggests someone who could make the introduction. If you just want to hire a software engineer, this isn’t that useful. But if you’re looking for something specific, like a growth person who’s done B2B at a company of a certain size, it really crushes.

Model Monday, Model Tuesday ... Models Everyday

We had about 40 power users at a Town event a couple of weeks ago and got a lot of product feedback. One person pointed out that during onboarding Town will say something like, “I’m going to create a routine.” Normal people trying Town for the first time don’t know what a routine is. Town should probably just say, “Do you want me to do this every day?” That's easy. You hear feedback like that and you fix it. Not all feedback is like that. My feeling is you should listen to all of it. You obviously can't implement all of it, but you can keep it in your head. Sometimes you find something you can build that solves a lot of the feedback at once. And then sometimes you hear the thing that everyone wants and you're still not convinced you should build it. Because if you go too far in that direction, the product kind of loses its soul. There's a Simpsons episode where Homer's brother decides Homer is the average American, so he should build the car Homer wants. He lets Homer design it and the car is a total monstrosity. That's kind of the problem. You have to listen, because if you don't listen you're going to build the wrong thing too. But at some point you still need intuition and taste about what makes the product simple and what will speak to lots of people. And if it works, people say you're a product genius because you made weird choices. If it doesn't work, they say you weren't. It's art more than science.

The vibe from my co-founder "i love how lazy this is, this is the way!" AI DOING WORK FOR YOU IS THE WAY!

One thing I like about Town is I get to do B2B sales some days and consumer marketing other days. They’re really different. B2B go-to-market isn’t the same for everybody, but the motions are similar. Consumer has many more things you could do, which means you have to be a little more picky about where you spend time and effort. It’s fun to get both problems at once.

We have tremendous product market fit with parents and families at Town. Schools send a ton of emails, there are portals for everything, sports, summer camps, scheduling, all this stuff you have to keep track of. Town happens to be really good at a lot of it because it's good at email and scheduling. The question for us is what to do with that. Parents are willing to pay and they talk about the product a lot in WhatsApp and Facebook groups. But the willingness to pay is obviously different from a mid-market company, and you don't get the same kind of expansion where you can land in one company and potentially spread across hundreds or thousands of people. Talked about it with Harry Stebbings on 20VC:

Unironically planning a trip with a personal assistant. @TownAI has got your back if you’re a working mom.

Humans and robots have one more thing in common: we love to talk talk talk! Talk with your Townie at town.com

I don’t really want to scroll through X every morning, so Town does it for me. We have a skill called X Daily Digest. You give it the accounts you want to track, it pulls the posts, ranks them and sends you a scannable digest. So basically every morning I get a summary of what’s happening from the people I care about on X. Then I use that to kick off ideas for things I could tweet as a follow up. This one is something anyone can install. It’s under Powers → Explore in Town.

Let’s go!

We're at a ~$75K/year run rate in AI tooling per engineer at Town right now. Here's the stack: - Devin for bugs, smaller changes and visual tweaks, largely because the team experience in Slack is great. - Cursor for more visual/front-end work. Composer is really fast. - Codex and Claude Code are roughly 50/50 for everything else. Five months ago I would've said we were mostly using Claude Code. But the new Codex is really good.