Platform Operations
What does it take to keep a platform running — and how do you make it better every cycle?
Platform operations is the discipline that sits between engineering and infrastructure. Engineering builds features. Operations ensures they deploy, run, and recover. Without operations discipline, every feature is a liability.
Operations Map
Three concerns, each with distinct ownership:
| Concern | Question | Where It Lives |
|---|---|---|
| Build | How does code become a deployed artifact? | CI Testing Infrastructure |
| Run | Where does it execute and how is it monitored? | Cloud Orchestration |
| Verify | How do we know it works and what it costs? | Testing Economics |
Decision Boundaries
Every platform has positions to fill. These are the active decisions:
| Decision | Question | Current Choice |
|---|---|---|
| Cloud Services | Where does it run? | Vercel (Web2) + Fleek (Web3) |
| Identity Stack | How do you prove who someone is? | Clerk + zkLogin |
| Data ORM | How do you talk to the database? | Drizzle |
| Message Bus | How do services communicate? | Kafka + Azure Service Bus |
| File Management | How do you manage files? | Evaluating |
| Payment Rails | How do you collect money? | Stripe + Crypto |
See Tech Stack for the full positions table and Tech Decisions for the evaluation process.
Dig Deeper
🗃️ Cloud
8 items
🗃️ Dev Ops
3 items
🗃️ Github
3 items
📄️ Linux
50 most used terminal commands
Context
- Platform — The ABCD thesis: AI + Blockchain + Crypto + DePIN
- Product Engineering — Workflow, quality gates, stack reference
- Tech Stack — Every position, every choice
- Standards — Where proven operations patterns graduate
Questions
What operations task are you doing manually today that should be automated infrastructure tomorrow?
- Which deployment failure mode has cost you the most time in the last month?
- If your CI pipeline disappeared, which quality gate would break first?
- What's the gap between your test environment and production — and what bugs live in that gap?