Hello, I'm
Bekir Can Yalcin
Computer Engineer
I build AI-driven applications and robotics systems that bridge research and production. Whether I'm developing autonomous delivery robots, architecting RAG-based document pipelines, or deploying embedded object detection for smart agriculture — I turn complex engineering challenges into working software.
02. Where I've Worked
Experience
- ▸Designed and developed a software pipeline utilizing RAG and NLP models to automate GTIP classification.
- ▸Developing an asset recovery system for idle stock and production surpluses.
- ▸Built an Allocated Lab Information Management system for intelligent document and user management.
- ▸Developed AI systems for R&D research.
- ▸Multi-angle color prediction model development based on pigment concentration.
- ▸Pigment concentration estimation from target optical values (inverse modeling).
- ▸Surface defect detection from microscopic images using YOLO.
- ▸Training object detection models and implementing them into embedded systems like Raspberry Pi.
- ▸Developing mobile applications and web pages using React and Node.js.
- ▸Developing AI-powered agricultural technology solutions, integrating IoT devices for precision farming.
03. Where I've Studied
Education
High School Diploma
Ilgin Science High School
04. What I Know
Skills
Languages
- Python
- JavaScript
- TypeScript
AI / ML
- PyTorch
- OpenCV
- Machine Learning
- Deep Learning
- RAG
- NLP
- YOLO
Web & Mobile
- React.js
- React Native
- Node.js
- MySQL
Systems & Tools
- ROS2
- 3D Modeling
- Raspberry Pi
- IoT
- Git
Languages
- TurkishNative
- EnglishB2
Certifications
- ▸Advanced Learning Algorithms — Andrew NG
- ▸Machine Learning Specialization — Andrew NG
- ▸Deep Learning Specialization — Andrew NG
05. What I've Built
Projects
Vision-Based Autonomous Delivery Robot
Final project for CENG483 Behavioral Robotics course at IZTECH. A monocular-camera-only indoor navigation robot that uses ArUco fiducial markers for target identification — no SLAM, no range sensors. A Finite-State Machine governs the mission: target search → visual alignment → PnP distance estimation → open-loop approach. Controlled via a lightweight web interface with live video feed and queue-based mission definition. Stack: ROS 2 Jazzy on Ubuntu 24.04, Raspberry Pi 5, OpenCV ArUco module, differential-drive 2WD chassis.
GTIP Classification Pipeline
A software pipeline utilizing RAG (Retrieval-Augmented Generation) and NLP models to automate GTIP (customs tariff) classification. Reduces manual classification effort significantly with AI-driven document understanding.
Multi-Angle Color Prediction Model
Deep learning model for predicting color appearance across multiple viewing angles based on pigment concentration. Also includes inverse modeling — estimating pigment concentration from target optical values.
Surface Defect Detection (YOLO)
Real-time surface defect detection system for microscopic images using YOLO object detection. Built for industrial quality control in R&D environments.
Precision Agriculture IoT System
AI-powered agricultural technology solution integrating IoT devices for precision farming. Includes sensor data pipelines, anomaly detection, and mobile dashboard for field monitoring.
Embedded Object Detection
End-to-end object detection pipeline deployed on Raspberry Pi. Trained custom models for real-world deployment on resource-constrained embedded hardware.
Lab Information Management System
Allocated Lab Information Management system with intelligent document processing and role-based user management. Built to streamline R&D lab workflows.
Graduation Management System
A web-based graduation management system that streamlines the graduation application process, document tracking, and approval workflows for university departments.
More projects coming as I continue building. See GitHub for the latest.
06. Get In Touch
Contact
I'm currently open to new opportunities — whether it's a full-time role, freelance project, or just a conversation about AI and engineering. My inbox is always open.
canyalcin111@gmail.com
linkedin.com/in/canyalcin111
GitHub
github.com/canyalcin1
YouTube
youtube.com/@canyalcin1