Overview
At Razor, I led the development of a high-volume pricing synchronization engine across Amazon, Walmart, Shopify, and Mercado Libre. Every channel had its own quirks — throttling, schema mismatches, and unpredictable response latency — yet the goal was simple: keep millions of SKUs in perfect sync, everywhere, always.
Problem
Razor’s catalog spanned multiple continents and currencies. A single missed update could cascade into revenue loss, pricing penalties, or buy-box drops. We needed a fault-tolerant system that could handle tens of thousands of concurrent updates per minute without breaching rate limits or data integrity.
Solution
I architected a distributed scheduler using async task queues, built idempotent API clients for each marketplace, and introduced an audit-first logging system that could replay failed updates deterministically. Every operation was version-tracked, timestamped, and recoverable — a design that transformed chaos into predictable state.
Impact
Automated pricing updates influenced $1M+ in monthly revenue
Achieved 99.98% update reliability across five marketplaces
Reduced manual price reconciliation time by 80%
Tech Stack
Python · FastAPI · AsyncIO · Redis Queues · PostgreSQL · Grafana · S3 Logging