.NET backend engineering from production work: high availability, adaptive testing, APIs and data pipelines. The systems behind these notes are written up as case studies, and the techniques as practical backend guides.
How an IRT-based adaptive test picks every next question: 1PL to 3PL models, EAP ability estimates, maximum information with exposure control and content balancing, and an API that stays responsive at exam peaks.
Mahdi Molavi
Read article
Video players send play, pause, seek and heartbeat events, and some of them never arrive. How I turned that noisy telemetry into reliable per-learner watch-time facts with .NET 10 and SQL Server.
Mahdi Molavi
Read article
Exam platforms idle for weeks, then everyone starts at once. Redis Sentinel failover, StackExchange.Redis configuration, stampede-safe caching and Locust load tests, from a setup that carried 45,000 to 70,000 concurrent users.
Mahdi Molavi
Read article
