At SONO YAZILIM A.Ş., I architected data analysis modules for a GRC enterprise system and optimized runtime performance by 15%. I also built responsive dashboards to display algorithmic metrics and predictive data.
At INOSAS, I developed and optimized a deepfake detection system using custom CNN architectures. I built a low-latency video preprocessing pipeline with MTCNN and used OpenCV to improve feature extraction speed under compute constraints.
I'm building an AI Compliance & Fraud Co-Pilot for SMEs, with a four-agent LangGraph pipeline that grounds audit reports in a RAG-based legal knowledge base. Its computer vision forensics module checks identity documents against Turkey’s KVKK data protection law.
My projects also include an ASL recognition system using TensorFlow and MediaPipe, and a medical assistant that retrieves information from medical references with RAG. I bring a Software Engineering background to developing and deploying machine learning systems.
