Day 1: AskedAgain
Most of Day 1 wasn't spent building. It was spent failing to pick an idea, until a scoped-down FAQ deduplicator finally shipped.
Day 0 was the starting line. No app had shipped yet. Day 1 was supposed to fix that, and it did, but not in the way I expected going in. Most of the day wasn't spent building anything. It was spent stuck on the part that actually turned out to be the hard part: picking what to build.
By 10pm I still didn't have an idea I felt good about. That's the honest version of Day 1, and it's worth writing down, because "I spent all day and almost didn't ship" is going to happen again, and I'd rather admit it now than pretend every day of this challenge goes smoothly.
The One I Didn't Build
For a while I was circling something like an SES console, but chat-based. Ask it what's happening with your email sending in plain language instead of digging through a dashboard. I liked the framing, but every time I tried to scope it down to something buildable in a day, it kept sprawling. A chat interface over infrastructure data is a real feature, and features want other features next to them, more data sources, more context, more edge cases in what people actually ask it. That's not what a 24-48 hour, no-database build is supposed to look like. I shelved it, not as a bad idea, just the wrong size for right now.
What Shipped: AskedAgain
What I built instead is called AskedAgain. You paste in a pile of questions you keep getting, out of DMs, emails, support tickets, wherever, messy and duplicated and out of order, and it comes back as a clean FAQ: near-duplicate phrasings merged, one canonical question per real topic, one answer each.
It's not a separate project. It lives at builduntilviral.com/askedagain, a new route inside this same app. No database, no accounts, no persistence. One page, one paste box, one request. You get an answer back, you copy it or share a link, and that's the whole interaction. Nothing about you or your questions sticks around on a server anywhere, because there isn't a server keeping anything.
That constraint wasn't a limitation I worked around, it's the reason the idea survived the same day the SES-console idea didn't. A stateless, single-request utility fits inside a day. A chat product with memory and context doesn't, not at this budget and not at this pace.
Why the Shape Matters More Than the Idea
The lesson from Day 1 isn't really about FAQs. It's that the constraint I set on Day 0, no database by default, isn't just about avoiding a free-tier bill. It's a filter for which ideas are even eligible to ship in a day. Anything that needs to remember something about a user across visits immediately costs more time than I have. Anything that answers one question and forgets you the moment you close the tab fits the shape of this challenge. AskedAgain fits. The chat console didn't, at least not yet.
What's Next
I'm still turning over the chat-based idea for later, if there's real interest in it. And I'm thinking about adding a simple way for people to drop suggestions for what to build next, since the whole point of shipping in public is finding out what people actually respond to, not guessing alone.
Follow the Challenge
If you want to follow along, head to builduntilviral.com and add your email or X handle to the wall of supporters. Email is optional, and you choose how often you hear from me: daily, weekly, monthly, or never. Only your X handle and avatar show up publicly on the wall.
Day 1 is live at /askedagain. See you on Day 2.