Why does AWS Lambda latency make my sourdough starter schedule impossible?

I am fed up with the way cloud services sabotage my baking rhythm. Every time I set a CloudWatch alarm to trigger a Lambda that notifies me when the starter reaches peak activity, the function fires 5‑10 minutes late. That delay means I miss the optimal window for feeding, and my dough collapses.

• EC2 instances spin up in 30 seconds, but the API Gateway I use for webhook callbacks takes another 45 seconds to become healthy. My 24‑hour proof cycle can't afford that jitter.
• S3 versioning is great until the eventual consistency delay shows me an outdated object, so I end up pulling the wrong hydration ratio.
• Terraform applies a new IAM policy, then the next run fails because the role propagation hasn't finished. I have to manually intervene, breaking the automation I built to free up time for kneading.

The whole stack feels like a fickle baker's assistant that forgets the oven is preheated. I need a cloud that respects the biology of yeast, not one that treats my schedule like a load‑balancer's whim.

Anyone else had their bread ruined by cloud latency? How do you force AWS to behave like a reliable kitchen timer?