Juggling 300 integrations while scaling to 10mil. Signals per day
summary
A team builds an F# service on GCP using serverless components (Cloud Functions / App Engine) to ingest and normalize over 10,000,000 GPS signals per day from 300 distinct providers. They ship a Fleet UI, a provider registry JSON, and a publish/subscribe API, use a provider integration template to onboard short-term devs, and tolerate a temporary shared-kernel for fast iteration before splitting contexts. Results: integrations and customers scale, but deployments and refactors force gradual contract codification and dedicated debt repayment.
AI-written summary · about
chapters (13)
- 0:00Introduction and team overview
- 1:06Project background and goals
- 10:43From infancy to childhood phase
- 13:54Choosing tech stack and architecture
- 18:10First customer and manual onboarding
- 20:39Team collaboration and async workflow
- 25:56Dynamic re-teaming and scaling challenges
- 29:47Bounded contexts and shared kernel issues
- 35:46Customer pain points and domain expertise
- 37:56Implicit modeling and technical debt
- 43:13Specialization and knowledge silos
- 46:26Key learnings and iterative process
- 50:40Closing remarks and Q&A
related talks
35mConvergence of Serverless APIs & Compute
40mMaking every minute count: How Nimses scales it's mobile platform on Google Cloud
27mScalable by Design: How to Build an Elastic Cloud Infrastructure...
44mScaling up a modern digital business to reach a million users
32mScaling Up for Scalable Apps: The CIF-N26 Experience
29mHow local businesses are rapidly achieving global scale on AWS
22mBlack Friday Brilliance: Managing a Billion Transactions with Tech, and Teamwork
51mMicroservices and Microfrontends: Scaling, Learning, Growing