METABYTE
Back to articles

Build Your Own Daily Notification Bot on AWS Free Tier: Weather, Wardrobe, and Zero Drama

Lambda checks the forecast, Bedrock tells you what to wear, and ntfy.sh pushes it to your phone — all on AWS Free Tier.

8 mai 20262 min read
Build Your Own Daily Notification Bot on AWS Free Tier: Weather, Wardrobe, and Zero Drama

Why agonize over what to wear each morning when you can offload that existential dread to a cloud AI? AWS proposes a bot that tells you whether to grab an umbrella or not. And yes, it's free — except for the few cents you'll spend on Bedrock API calls.

How It Works

The architecture is simpler than your morning coffee routine:

  • EventBridge triggers a Lambda once a day (say, 7 AM).
  • Lambda fetches the weather forecast from a public API.
  • Bedrock (with Claude or Llama) analyzes the data and spits out a recommendation like "It's 15°C and rainy — wear a raincoat and bring an umbrella."
  • The result is pushed to ntfy.sh, a free push notification service that delivers it to your phone.

Everything fits within AWS Free Tier (1M Lambda requests/month, 1K Bedrock requests). The only risk is forgetting to shut down an instance, but that's on you.

Why It's Cool (and Slightly Painful)

On one hand, it's a brilliant serverless architecture: minimal code, maximum automation. On the other, it's like assembling IKEA furniture without instructions — all the pieces are there, but good luck figuring out how they fit. If you have an hour to spare and a desire to play with AWS, this is the perfect pet project.

Bonus: You can adapt it for anything — meeting reminders, CI/CD status, or even "feed the cat" (if you add an IoT feeder). Just don't forget to subscribe to your own ntfy channel, or your notifications will vanish into the void, like your plans for a full rewrite.

METABYTE studio comment: This bot is a perfect sidekick for any microservice — be it a deploy reminder or a habit tracker. Just make sure you actually subscribe to your own ntfy topic, otherwise your notifications will end up in the same place as your abandoned side projects.

NEXT STEP

Liked the approach?

We apply the same principles to client projects: AI, automation, products that don't die after launch.