Why I'm Building a $10 Pager

Here's a business plan that sounds stupid: build an incident management tool as a solo developer from northern Minnesota, competing with established players that have hundreds of employees and fifteen years of head start.

Charge a tenth of the price. Keep it simple.

That's Siren. Let me explain why I think it's not as stupid as it sounds.

---

The Premise

I'm coming from Mega-Corp. My time wasn't always wasted there—but when it started being wasted, and I saw what AI made possible, the contrast became unforgivable. When I started to see what 10x really meant for me—not a buzzword, but actually shipping in hours what used to take days—I couldn't tolerate features taking a quarter. Tickets rotting in backlogs. Finally finishing something only to hear "that's great, but that's not our area."

Continuing felt like betraying myself.

I had to find a way to use these new tools. Siren is my "it doesn't have to be this way." I'm building a $10 pager because I'm capable of building an enterprise-grade pager at a price that's difficult to compete with.

---

The Math That Changed

Here's what's different now versus five years ago.

Managed infrastructure is stupid good. Siren runs on CockroachDB—distributed SQL, multi-region replication, automatic failover, managed backups. Netflix runs their production on it. I get the same infrastructure on a free tier. No DBA required.

Redundancy is cheap. Siren pages through both Twilio and Vonage. If Twilio has an outage—and they do—Vonage keeps paging. Dual-provider messaging used to be "enterprise architecture." Now it's two API keys and some routing logic.

AI makes one developer dangerous. I've been doing this for fifteen years. IT support, dev, SRE, infrastructure—small shops where you see everything, big orgs where you learn the hard way why things are done the way they're done. You sit in enough war rooms, debug enough production incidents, listen to enough smart people explain the thing you don't understand yet, and eventually you've seen a lot of tools.

Then AI comes along and says: "All that stuff you know how to do? You're an expert at that now."

Not in the "AI writes my code" way that produces garbage. In the "I move 10x faster on things I already understand" way. Claude doesn't replace fifteen years of carrying pagers and debugging production at 3am. It multiplies it.

---

Why Not Build It Yourself?

If the math changed, why not do it yourself?

Honestly, maybe you should. A lot of you could. If you work 80 hours a week and already know exactly what to do, you'll have something like Siren in about a month.

Then you'll have to maintain it. Keep it secure. Fix the bugs. And if you want it actually redundant? Auto-failover when Cloudflare has a bad day, multi-cloud, geo-distributed database? That's not a month anymore.

If you're here, you already have a product that takes your time. I'm trying to build something you can set up in five minutes that costs so little you don't even consider building it yourself.

The decision should be obvious: "I could build this, but why would I bother for $10/month?"

That's the spot I'm aiming for.

---

The Bet

The economics work. $10/month for 50 incidents. Unlimited seats. No per-user pricing. I can sustain that because my costs are low and I don't have investors demanding growth at all costs.

Small is an advantage. I ship faster than an enterprise product team can schedule a meeting. No legacy code, no legacy customers demanding features. I can make opinionated choices and stick with them.

Maybe I'm wrong. Maybe there's something about incident management at scale that I'll only learn when I hit it. Maybe the incumbents' moat is deeper than it looks.

But I've been building software long enough to bet on simple and reliable over complex and feature-rich.

---

What Siren Actually Is

You get an email address. Point your monitoring at it—Datadog, Grafana, CloudWatch, a cron job, whatever. When an alert comes in, Siren pages whoever's on call. If they don't answer, it tries again. Then it escalates, tries that person once more, and stops. After that you're on your own—I tried to wake up your team.

This limitation keeps the cost low.

You set up your rotation. Daily or weekly. Override it when someone's on vacation.

That's it.

$10/month. 50 incidents included. Unlimited team members. No seat licenses, ever.

25 incidents free to start. No credit card.

---

The Invitation

I'm not going to pretend Siren is proven. It's new. I'm one guy. You've never heard of Split Rock AI and you probably can't find Grand Rapids, Minnesota on a map—though Bill Baker scored an important goal in the 1980 Olympics and he's from here, so you should know that.

But the product works. The infrastructure is solid. The price is right.

If you're a small team paying enterprise prices for a pager you don't fully use, give it a look. Try the free tier. See if it does what you need.

And if it doesn't—tell me why. The roadmap isn't set by a product committee. It's set by what actually matters to the people using it.