I ship backends fast — and keep them reliable under real users.
Early-stage products need speed and survivability — clean architecture, observability, safe rollouts, and guardrails that prevent costly mistakes as you grow.
Talk: https://cal.com/miladtsx/intro
- Block By Block — community growth + engagement tools
- CommunityPulse — community steering system (measure → recommend actions → improve)
- Predictable Job Queue (side project) — retries, idempotency, and recovery boundaries in distributed systems
- You want one engineer to own the system from
idea → production → calm operations - You need an MVP shipped quickly, but you can’t afford outages, data corruption, or runaway costs
- You hit first traction and need the system to scale without becoming fragile
Operating principles
- Start from product risk: what would kill growth or trust?
- Ship the smallest version that is observable and recoverable
- Add guardrails early (validation, rate limits, idempotency, safe retries)
- Prefer boring designs that scale predictably
Results
- Runway: reduced cloud spend ~60–70%
- Stability: improved crash-free sessions ~65% → ~92%
- Scale: grew real-time backends ~100 → 2,000+ concurrent users
- Launch: shipped investor-ready MVPs in days/weeks