System Design
Multi-Region Service Boundaries
Partition workloads, latency budgets, and ownership lines when your product spans more than one geography.
Overview
This cohort walks through request paths, data proximity, and release cadences for teams shipping one codebase to several regions. You leave with diagrams you can paste into internal docs and a checklist for operational reviews before you expand another cell.
Included
- Boundary exercises using anonymized SaaS traffic shapes
- Read path vs write path worksheets with facilitator markup
- Runbook skeletons for regional degradation drills
- Pair sessions on dependency graphs between services
- Async review of your current topology (one submission per team)
- Office hours on configuration drift between environments
- Lightweight templates for cross-team RFC comments
Outcomes
- Document a north-star topology your leads can defend in planning forums.
- List three concrete rollback gates tied to regional deploys.
- Name the single owner for each externally visible endpoint slice.
Lead contact
Haneul Park
Former platform lead for a Seoul-based logistics SaaS; now coaches architecture cohorts across APAC.
FAQ
Do you cover specific cloud vendors?
We stay vendor-neutral on primitives (queues, object stores, DNS). Bring your stack notes and we map patterns to what you already run.
Is there homework between sessions?
Yes—short written artifacts (usually two pages) so instructors can comment asynchronously. Expect roughly four hours weekly.
What is not included?
We do not perform production changes on your systems, and we do not certify vendors. The work is educational and advisory.
Recent notes
“Week four’s boundary tracing exercise finally made our shadow traffic story legible to product. I still wish we had one more hour on cache stampede edge cases.”
“Templates were blunt in a good way—no decorative boxes, just prompts that forced crisp decisions.”