↓Skip to main content

Blog

Detailed engineering write-ups on the systems I have designed and operated. Each study focuses on the decisions, operating constraints, and lessons that shaped the work.

Scaling the Invoice Pipeline 20x
·2767 words·13 mins
Case-Study Distributed-Systems Payments Reliability Scaling
What changes in the invoice pull-and-present design at 100K and 1M invoices a day - architecture options from scheduled Spot to serverless and managed orchestration, plus the partner rate limit and notification bill.
Invoice Presentation Orchestration
·2884 words·14 mins
Case-Study Distributed-Systems Payments Reliability
How an asynchronous, stateful workflow supported 50K-60K invoices per day across roughly 400 partner systems.
Scaling webhooks 10x: an ideal design
·1656 words·8 mins
Event-Driven Webhooks Serverless Platform-Engineering Scalability
The current webhook design holds ~100K events a day with six workers. Ten times that, with unpredictable spikes, needs a different shape — transactional outbox, event bus, serverless fan-out and delivery, and per-partner isolation.
Webhooks Platform
·3424 words·17 mins
Event-Driven Webhooks Platform-Engineering Reliability
How an event-driven webhook service delivered ~100K events a day to roughly 400 partner systems with at-least-once delivery, bounded retries, and a failure trail both sides could inspect.