Name: Baha Erdoğan
Email: [email protected]
Youtube: youtube.com/bahaerdoğan
Linkedin: linkedin.com/in/bahaerdogan/
I am a Computer Engineer specializing in Artificial Intelligence and Computer Vision at Huawei R&D. Driven by relentless curiosity, I focus on building high-performance machine learning models, autonomous systems, and scalable software architecture.
Beyond core engineering, I publish technical essays analyzing AI developments, computer vision methods, economic value dynamics, and historical perspectives. Having traveled across 15+ countries to observe technological infrastructure firsthand, I bring a multi-disciplinary approach to engineering and problem-solving.
Deep dives into AI architecture, computer vision algorithms, and socio-economic essays.
Data is an integral component in various aspects of our lives, from communication backbones to sophisticated artificial intelligence models that shape our future. Explore what makes data valuable and how modern systems leverage it.
Agentic AI and autonomous AI agents are revolutionizing software capabilities. This article breaks down how agentic architectures operate, moving beyond passive text generation to autonomous task execution.
Defense systems represent some of the most robust software architectures. An analysis of layered security, computer vision integration, autonomous defense protocols, and vulnerability dynamics.
Engineering Notes, AI Research, History & Technology Insights
Explore the legendary Turkish epic of Oğuz Kağan, a tale of leadership, conquest, and societal foundation. Discover how this ancient narrative influences cultural history.
A comprehensive guide to understanding machine learning fundamentals, algorithm classifications, optimization objectives, and real-world deployment patterns.
Delve into the Ergenekon legend, a foundational narrative of resilience and cultural rebirth. Learn about its historical context and historical interpretations.
An inquiry into productivity metrics, AI-assisted economic leverage, and how automated tooling shifts professional value attribution across technical fields.
Unlike standard supervised learning tasks, anomaly detection operates on severe class imbalance. A review of statistical, distance, and deep learning outlier detection techniques.
A technical breakdown of estimating object velocity in uncalibrated monocular video streams using Structure from Motion (SfM) and 3D camera reconstruction algorithms.
Explore articles on AI, Technology, History & Engineering
Engineering R&D and Software Industry Track Record
Current | Istanbul, Turkey
Artificial Intelligence R&D Specialist — Engineering state-of-the-art computer vision models, deep learning pipelines, and optimized inference algorithms for high-throughput visual processing.
2023 | Riga, Latvia
Deep Learning Research Program — Conducted robotic vision research and neural control applications (Baxter Robot Revival via Deep Learning Applications).
2024 | Istanbul, Turkey
Software Engineer (Data Team) — Built data processing pipelines, ETL tools, and analytical reporting features.
2024 | Istanbul, Turkey
Software Engineer (Profitability & Analytics Dept) — Engineed database algorithms and financial analytical software modules.
2023 | Istanbul, Turkey
Center of Excellence Department — Analyzed software quality assurance lifecycles, automated testing infrastructure, and AI-driven testing tools.
2022 | Antalya, Turkey
Full-Stack Web Development — Developed responsive Web applications using modern JavaScript, ReactJS, Node.js, and RESTful APIs.
Production Architectures, Engineering Trade-Offs & Systems
2024
Problem: Processing dense technical documents causes reader fatigue and delays executive decision-making.
Architecture: Built a browser-side document parser paired with light-weight NLP abstraction models to extract key entity graphs and semantic summaries.
Trade-off: Prioritized client-side zero-latency rendering over cloud-heavy LLM pipelines, preserving user document privacy while drastically cutting infrastructure overhead.
2022
Problem: Centralized file transfers over local networks suffer from bandwidth bottlenecks and single-point server failures.
Architecture: Engineered a custom multi-threaded Python TCP/UDP socket network protocol featuring asynchronous chunk verification and direct peer discovery.
Trade-off: Accepted initial peer discovery latency to achieve maximum throughput across local subnet links without reliance on external cloud coordination.
2023
Problem: Standard Python-based recommendation scripts fail to scale under high-concurrency matrix operations.
Architecture: Developed a bare-metal C++ recommendation engine utilizing optimized sparse matrix representations and parallel gradient descent for low-latency scoring.
Trade-off: Sacrificed rapid prototyping speed for sub-millisecond execution times and minimal memory footprints under large item sets.
2022
Problem: Standard touch user interfaces are inaccessible for children with severe motor skill impairments.
Architecture: Designed a native Android application integrating real-time gaze estimation algorithms using mobile front-facing camera frames.
Trade-off: Balanced frame rate sampling against battery thermal throttling to maintain continuous 30fps tracking on consumer mobile hardware.
Core Technical Stack & Engineering Competencies
Academic Foundation & Leadership Programs
2020 – 2025
Completed Computer Engineering degree with Full High-Merit YÖK Scholarship.
2022 – 2025
Pursued minor degree in Business Administration, studying financial analysis, management, and tech economics.
2024 | Riga, Latvia
Selected as Turkish delegate for international engineering and leadership initiative.
Social Responsibility and Educational Outreach
2020 – Present
Active community volunteer organizing educational and youth development projects.
2021 – Present
Active volunteer for environmental conservation and reforestation programs.
2020 – Present
Creating educational video content sharing software engineering principles, AI tutorials, and career insights.