This is an excellent story; thanks for sharing.
One thing I keep wondering: can you speak to the business implications of this? I love the idea of serverless, but it seems you’re entirely at your vendor’s mercy now from an operations standpoint. And I’m not necessarily thinking about downtime, because that happens. But if AWS says “you know what? Lambda was nice but we decided to sunset it”, you’re pretty much forced into a potential large-scale re-write on a short timeframe, no?
I guess one mitigating factor there would be that it’s easier to go from micro → monolith than the reverse. But still, curious on your thoughts on this.