At Avije Pardaz Pars, I architected and implemented a Django backend platform spanning roughly 40,000 lines of production code, from system architecture and database schemas to REST APIs and deployment.
I optimized Django ORM queries to reduce database load by about 40%, maintaining sub-200 ms response times for 1 million daily requests.
I also designed and deployed a real-time license plate detection and OCR pipeline with YOLOv11 and ONNX Runtime, processing more than 50,000 detections daily with end-to-end latency below 400 ms.
Alongside backend and machine learning work, I’ve built authenticated REST APIs for freelance clients, collaborated on Golang and Rust services, and support IT systems operations at GWDG.

