Why I Built Pingfyr

I needed to do something later.

That's the problem I kept running into — across fifteen years in digital products, my own projects, two companies (automatics.dev and optimotion.dev), and a growing obsession with AI agents. Different contexts, same wall.

Send a message in two hours. Fire a webhook three days after a user signs up. Wake an AI agent at a specific time to finish what it started. Remind someone before a deadline. The trigger wasn't complicated. The timing was the whole point.

The first few times, I reached for the obvious solutions. Cron jobs — the classic move. Set it up, deploy, forget about it. Except you can't really forget about it. Cron has no retry logic. No delivery receipts. If the job fails, you probably won't know until a user asks why they never got the email they were expecting. I've been there more than once.

So I tried AWS EventBridge. It works — it's built for this — but for what I needed, it felt like renting a warehouse to store a single box. The configuration alone took an afternoon. IAM roles, event buses, targets, DLQs. By the time I had it working, I'd forgotten what I was originally building.

What surprised me was that this wasn't just my problem.

At automatics.dev and optimotion.dev, I spend a lot of time in the weeds with businesses — understanding their workflows, building systems that connect things together. And in those conversations, the same request kept coming up, just in different words. A team needed to send follow-up messages to prospects after a specific number of days. Another needed to trigger the next step in a process hours after the first one completed. Others wanted alerts fired on a schedule, or reminders sent before something expired. Different industries, different contexts — same underlying need.

Do something. At a specific time. Reliably. Without building an entire infrastructure project to make it happen.

So I built Pingfyr. Not a glamorous origin story — just the cumulative frustration of knowing this should already exist and it doesn't, at least not in a way that doesn't require a week of setup for something that should take five minutes.

One API call: what to send, where to send it, when to send it. Email, Slack, Discord, Telegram, webhooks, AI agents. Retries and delivery logs included. You don't have to think about any of that.

I built it for myself. Turns out a lot of people need the same thing — developers, founders who just want the automation to run without becoming an infrastructure project, operations teams, AI builders trying to make agents that can actually wait for something.

If you've ever written a cron job for something that felt too simple to deserve a cron job, you'll get it.

Pingfyr is live — try it free at pingfyr.com. No credit card required.