I'm Nashit, an AI engineer and researcher based in Türkiye.
I build vision and deep learning systems, from research to production.
Computer vision and deep learning, from research to production
Peer-reviewed work and papers currently in preparation
From first prototype to a model running in production
Detection, tracking and stereo-vision pipelines with YOLO11, ByteTrack and OpenCV.
CNNs, UNet, LSTM and TSMixer models built, evaluated and compared on your own data.
Python services containerised for on-prem GPU hardware or the cloud, with GitHub Actions CI/CD.
Experiment design, evaluation and write-ups, from grant-funded projects to journal papers.
A short timeline of research and engineering
Led engineering at an AI automation studio serving clients in the US, UK and Pakistan, delivering voice AI, RAG systems, computer vision pipelines and workflow automation.
Quality assurance and technical report review on the eluna industrial software platform.
Configured and maintained FBB-IP network equipment and troubleshot connectivity issues in a live telecom environment.
Thesis on hybrid AI architectures for smart greenhouses, within the TÜBİTAK 1071 bilateral R&D project. Advisor: Prof. Dr. Yusuf Sinan Akgül.
Final-year project on AI-based recognition and synchronisation of Quranic text with audio. Applied ML projects in medical imaging, including knee osteoarthritis grading and melanoma detection.
I'm an AI engineer and researcher working on computer vision and deep learning for smart greenhouses, as part of an MS in Artificial Intelligence at Gebze Technical University. I studied Electrical Engineering at UET Taxila, and I've led AI and backend engineering at SoftDev, an automation studio serving clients in the US, UK and Pakistan.
Get in touch
Your answers await right here
Yes. I take on a limited number of computer vision and deep learning projects alongside my MS research. Send me a short description and we can find out if it's a good fit.
PyTorch and TensorFlow for training, OpenCV for classical vision and stereo geometry, YOLO11 with ByteTrack for detection and tracking, UNet for segmentation, and FastAPI with Docker for serving.
Yes. My main research is computer vision for greenhouse crops: fruit sizing from stereo imagery, water-stress sensing from RGB cameras and yield estimation, supported by a TÜBİTAK-funded project.
Yes. I've packaged vision pipelines as Python services for on-prem GPU hardware, reading RTSP camera streams and storing results in S3-compatible storage, shipped with Docker, Kubernetes and CI/CD.