Sign in Subscribe

My Portfolio

Experienced in developing and deploying high-performance ML models using Python, PyTorch, and TensorFlow.
Specialized in MLOps (CI/CD pipelines, monitoring, lifecycle management) and production deployments with PostgreSQL, Docker, Apache tools (Airflow, Spark), and AWS.

LinkedIn | GitHub | Medium | Google Scholar | hamzagorgulu7@gmail.com

Professional Experience

Machine Learning Engineer – HUB Enerji
March 2024 – Present | Istanbul, Turkey

  • Designed, developed, and deployed advanced forecasting models for electricity markets.
  • Built scalable, efficient solutions for market participants.
  • Achieved best MCP forecasting performance in Turkey.
    Tech: PyTorch, Scikit-Learn, GitHub Actions, PostgreSQL, Docker, MLOps, Apache Airflow/Flink/Spark, AWS, Databricks, Optuna, MLflow.

Computer Vision Engineer (Full-time Freelance) – Desion GmbH
Nov 2023 – Mar 2024 | Darmstadt, Germany

  • Improved model performance by 4% via industry-aligned adjustments.
  • Developed conveyor belt ticket defect detection script for production.
  • Automated labeling issue detector, increasing mean IoU by 6%.
    Tech: OpenCV, PyTorch, TensorFlow, Linux, Git, Docker, Cleanlab.

AI Researcher – Koç University & İşbank AI Center
Feb 2022 – Nov 2023 | Istanbul, Turkey

  • Developed novel self-supervised deep learning architecture for industrial FDD.
  • Conducted TÜPRAŞ-funded research integrating computer vision and NLP.
    Tech: PyTorch, TensorFlow, Docker, Git, SQL, Linux.

Projects & Publications

Small Object Detection with YOLOv5 – 2023
GitHub
Enhanced YOLOv5 with residual connections for small object detection.
Trained on custom dataset, deployed on Nvidia Xavier, +5% performance boost via pre/post-processing.
Tech: Python, PyTorch, YOLOv5, Nvidia Xavier, Docker, TensorFlow, ONNX.

Machine Learning Methods for Alarm Prediction in Industrial Informatics – 2023
Springer Link | Drive
Compared LSTM and Transformer models for industrial alarm prediction.
Provided actionable insights for improving alarm system accuracy.
Tech: Python, TensorFlow, LSTM, Transformer, Pandas, Matplotlib.

Education

Koç University – MSc, Computer Science & Engineering (2022–2024)
Focus: Self-supervised Learning, Fault Detection & Diagnosis, Computer Vision, NLP
TA Roles: Python Programming (COMP125), Summer Practice (COMP391)

Technische Universität Dortmund – Erasmus, Automation & Robotics (2018–2019)

Yıldız Technical University – BSc, Electrical Engineering (2016–2021)

Skills

Technologies: Python, PyTorch, TensorFlow, Transformers, Ray, Optuna, OpenCV, FastAPI, Docker, Kubernetes, Airflow, MLflow, PostgreSQL, MongoDB, InfluxDB, Redis, Linux, Git, GitHub Actions
Languages: Turkish (Native), English (C1), German (A2)

Contact

📧 hamzagorgulu7@gmail.com
🔗 LinkedIn
🔗 GitHub
🔗 Medium
🔗 Google Scholar