Voluntary Experience

cat ~/.voluntary

Software Engineer Intern

@ Technology Development Group
Mar 2023 — Apr 2023 · Türkiye
  • ▸ Assisted in developing a fully-autonomous cargo carrier vehicle that follows specified strips using ROS2 Humble and Python.
  • ▸ Built a graphical user interface with Python and Tkinter, forwarding sensor data directly to the user in real time.
Python ROS2 Tkinter

ML Engineer Contributor

@ UnifyAI
Jun 2022 — Jul 2022 · Remote
  • ▸ Contributed to the documentation of a large international open-source ML framework by testing functions and optimizing performance using Python and Git.
Python Git Machine Learning

ML Engineer Freelancer

@ Nutzentech
Sep 2021 — Jun 2022 · Türkiye
  • ▸ Built object detection models to detect surface damage in solar panels and wind turbines using PyTorch, TensorFlow, and AlexNet as the CNN architecture; trained and deployed on AWS SageMaker.
  • ▸ Analyzed air pollution data from drone surveys around factories, visualized hazardous zones on maps, and delivered reports with proposed pollution-reduction solutions.
Python PyTorch TensorFlow AWS SageMaker Computer Vision

AI Engineer Intern

@ Apziva
Dec 2020 — Mar 2021 · Remote
  • ▸ Built an XGBoost classifier using applicant data to predict candidates likely to submit applications; deployed with serverless AWS SageMaker Inference, reducing HR candidate-tracking costs by 30%.
Python XGBoost AWS SageMaker Machine Learning

R&D ML Engineer

@ CENGA R&D Team – Çukurova University
May 2020 — Apr 2021 · Adana, Türkiye
  • ▸ Collaborated on a deep learning UI for regression, classification, and time series forecasting—architectures including RNN, MLP, SVM, Random Forest, LGBM, CatBoost, XGBoost, Ridge, SARIMA, ELM, and more, built with TensorFlow and Scikit-Learn.
  • ▸ Contributed by collecting and summarizing research papers and patents, synthesizing concepts into new approaches and project ideas.
Python TensorFlow Scikit-Learn Deep Learning

Computer Vision Engineer

@ Çukurova Hydromobile Team
Sep 2019 — Oct 2020 · Adana, Türkiye
  • ▸ Built line detection algorithms using Hough transform and improved robustness with grayscale conversion and image sharpening.
  • ▸ Trained object detection models (DarkNet YOLOv4) to detect traffic signs and lights in varied lighting, achieving 90% accuracy with a manually curated dataset.
  • ▸ Integrated line detection and object detection on Nvidia Jetson Nano using Jetson Inference—team placed 2nd in the 2020 TÜBİTAK Efficiency Challenge.
Python YOLOv4 OpenCV Nvidia Jetson Computer Vision