Aysima merve (Merve) A.
AI & Computer Vision Engineer (Internship & Thesis)
Experience
Nov 2024 - May 2025
7 monthsPadua, Italy
AI & Computer Vision Engineer (Internship & Thesis)
Codebeex SRL
- Boosted vehicle image processing speeds by 99.6% by architecting a scalable AI backend with FastAPI microservices.
- Improved car part segmentation mAP@50-95 by 32.6% by creating and training on a bespoke, data-centric dataset.
- Slashed inference latency by up to 66% (from 1.5s to 0.5s) via an in-memory caching strategy.
- Reduced CPU inference time by over 60% by converting PyTorch models to the ONNX format for deployment.
- Engineered a novel damage assessment system with 99% accuracy in flagging reusable components.
- Engineered a privacy-preserving module using Florence-2 (OCR) and custom-trained YOLOv11.
Dec 2019 - Aug 2021
1 year 9 monthsAnkara, Turkey
BIM Modeler - 3D Modeler
Korfez Engineering
- Generated CityGML-compliant 3D building models for 1000+ structures in Izmir using MicroStation.
- Collaborated with engineers to meticulously validate architectural drawings, ensuring high model consistency and geometric accuracy.
Sep 2019 - Jun 2020
10 monthsAnkara, Turkey
Geomatics Engineering Research
Hacettepe University
- Reduced manual reporting time by approx. 3 hours per project by developing an automated web application.
- Produced LiDAR-based photogrammetric maps for structural monitoring.
Summary
Computer Vision Engineer specializing in production-ready AI systems. Expert in segmentation, OCR, and model optimization with PyTorch, ONNX, and FastAPI. Delivered key improvements including +32.6%mAP, -66% latency, and -60% CPU time in automotive e-commerce applications.
Skills
- Programming & Scripting: Python, Matlab, Sql (Postgresql, Mysql), Javascript, Html, Css, Uml, Vs Code, Jupyter Notebook
- Ai & Computer Vision Frameworks: Pytorch, Tensorflow, Keras, Scikit-learn, Opencv, Detectron2, Albumentations, Kornia, Transformers (Hugging Face), Langchain, Llm, Rag, Openai Api, Cursor, Numpy, Pandas, Torchvision, Pil
- Specialized Models & Tools: Yolo (V8, V11), Sam, Florence-2, Paddleocr, Tesseract, Cvat, Roboflow
- Geospatial & Mapping Tools: Arcgis, Qgis, Erdas, Agisoft, Postgis, Geospatial Data & Processing
- Deployment & Mlops: Fastapi, Docker, Onnx, Torchscript, Openvino, Git, Github, Gitlab, Uvicorn, Rest Apis, Faiss, Tensorboard
- Technical Concepts: Deep Learning, Machine Learning, Computer Vision, Model Optimization, Real-time Inference, Object Detection, Image Segmentation, Ocr, Nlp, Data-centric Ai
Languages
Turkish
NativeEnglish
AdvancedItalian
ElementaryEducation
Sep 2022 - Jul 2025
University of Padua
Master of Science, ICT for Internet and Multimedia · ICT for Internet and Multimedia · Padua, Italy
Sep 2016 - Jun 2020
Hacettepe University
Bachelor of Science, Geomatics Engineering · Geomatics Engineering · Ankara, Turkey
Need a freelancer? Find your match in seconds.
Try FRATCH GPT More actions
Similar Freelancers
Discover other experts with similar qualifications and experience