Engineering Journey & Academic Background
My name is Ataberk Susam. I am currently studying Mechanical Engineering at Middle East Technical University (METU) in Ankara, Turkey. The rigorous analytical foundation provided by mechanical engineering—spanning numerical methods, kinematics, signal analysis, thermodynamics, and physical modeling—heavily shapes my approach to software engineering and system architecture.
Rather than viewing software as abstract code, I treat computational tools as precision mechanical systems: workflows must minimize friction, resource allocation must avoid thermal and memory bottlenecks, and data flow pipelines must operate deterministically with verifiable reliability.
"Engineering is the art of eliminating unnecessary overhead." In software, that means running operations on the client when possible to eliminate latency, respecting user privacy by never collecting unnecessary data, and engineering lightweight architectures that execute in milliseconds rather than seconds.
Core Technical Specializations
Over the past several years, my development work has focused on three primary interconnected disciplines:
1. Real-Time Computer Vision & Image Processing
Leveraging OpenCV, Tesseract OCR, and custom image segmentation pipelines, I build vision tools capable of interpreting complex real-world and screen data in real-time. This includes projects like Okey Helper, which processes video feeds at low latency to extract board state information, classify tiles, and compute strategic moves on the fly.
2. Zero-Upload, Privacy-Preserving Web Tools
A core tenet of modern computing is that sensitive user data should never leave the user's local environment unless strictly necessary. To demonstrate this principle, I created MassForge PDF—a browser-based PDF suite powered entirely by client-side JavaScript (pdf-lib, pdf.js). Users can merge, split, encrypt, watermark, and compress documents without a single byte ever being transmitted to a remote cloud server.
3. High-Performance Desktop GUIs & Automation
For data-intensive, multi-threaded tasks that demand native execution, I engineer desktop applications using Python, PyQt6, and CustomTkinter. Projects like KLScrapper utilize asynchronous Playwright browser rendering alongside FFmpeg stream muxing to download and process high-resolution media streams reliably.
Technical Skills & Tooling Matrix
| Category | Primary Technologies & Frameworks | Applied In |
|---|---|---|
| Languages | Python 3.x, JavaScript (ES6+), HTML5, CSS3, SQL, Shell | All Software Systems |
| Computer Vision | OpenCV, Tesseract OCR, NumPy, Color Space Masking, Contour Analysis | Okey Helper, Screen Automation Tools |
| Client-Side Web | Vanilla JS, pdf-lib, pdf.js, Web Workers, Canvas API, ArrayBuffers | MassForge PDF Web Suite |
| Desktop GUI Frameworks | PyQt6, CustomTkinter, Qt Designer, PyInstaller | KLScrapper, MassForge Desktop Tools |
| Automation & Scraping | Playwright, FFmpeg, Requests, BeautifulSoup4, Multiprocessing | Media Crawlers, Stream Downloaders |
| Version Control & DevOps | Git, GitHub, GitHub Actions, Linux Environments | CI/CD, Open-Source Repositories |
Open-Source Commitment & Community Contributions
I strongly believe in the open-source ethos. Transparent code builds trust, enables peer review, and allows other developers to learn and contribute. Most of my desktop tools and scripts are open-sourced under permissive licenses (such as the MIT License) on my GitHub profile (@klausimon).
Whether you are interested in discussing computer vision pipelines, client-side web architectures, Python desktop development, or potential project collaborations, I welcome communication. Visit the Contact Page or reach out on GitHub.