An internship focused on making an existing enterprise MRP system faster and more reliable. I learned that performance work is mostly careful measurement — finding the real bottleneck before touching a line of code.
2x
Workload capacity
12+
Production APIs
99.95%
Uptime
01
What I did
Profile and optimize database operations in an enterprise MRP system
Automate background tasks to handle higher load
Build and document production-grade APIs
02
Key wins
01
Optimized MRP system to handle 2x higher workload by streamlining database operations and automating background tasks
02
Built 12+ production-grade APIs sustaining 99.95% uptime
03
Work & life
Measure, then optimize
Doubling throughput sounded like a rewrite at first — it turned out to be a handful of well-chosen indexes and moving slow work off the request path. A good lesson in not guessing.