Omer Sukru Varol

I’m an R&D Engineer specializing in Industrial AI, computer vision, and embedded systems. I focus on deploying production-ready ML solutions with measurable performance—sub-150ms latency and 95%+ accuracy—while keeping architectures cloud-native and maintainable.

At FarmaKod, I architected acoustic resonance NDT production-line systems using pneumatic excitation and IEPE sensors. I built a real-time FFT/STFT signal processing pipeline and developed a PyTorch defect detection model, delivering <150ms source-to-decision latency with 95%+ accuracy on 10k+ test samples.

I also deliver edge AI solutions by deploying YOLO-based object detection/tracking on Jetson Nano and Raspberry Pi. My work includes managing a dataset pipeline for 20k+ annotated images to achieve ~40 FPS on embedded hardware, as well as sensor-fusion localization using GPS/IMU with ±2cm positioning accuracy.

Beyond core ML and sensing, I’ve led cloud modernization by migrating a PHP platform to GCP using Cloud Run and Cloud SQL, improving uptime to 99.9% and reducing infrastructure costs by 35%. I’m energized by end-to-end ownership—from C# .NET/MSSQL backend development and RESTful APIs to building a solo-developed iOS language learning app with Claude API, Firebase, and RevenueCat.