Software Engineer — AI, CV, Embedded Systems at Mersad AI (2025-12 – 2026-05)
Engaged in the design, integration, and development of electronic and software components for next-generation smart infantry systems and Unmanned Combat Aerial Vehicle (UCAV) projects at a leading defense company based in Doha, Qatar.
- Led the software and electronics development of the project from initial design through deployment, including hardware architecture decisions, component selection, PCB-related circuit design considerations, and production-oriented system integration.
- Designed, trained, converted and optimized YOLO object detection models using datasets containing over 30,000 images, optimizing model accuracy and inference efficiency for deployment on low-power edge AI platforms equipped with 3–6 TOPS NPUs.
- Designed and implemented embedded electronic subsystems, including voltage dividers, MOSFET gate drivers, LED driver circuits, and supporting control electronics for Edge AI-based Infantry Equipment platforms.
- Developed firmware for an auxiliary microcontroller responsible for driving high-lumen LED systems, managing Long Range Acoustic Device controls, and providing communication interfaces with the SBC running onboard AI inference workloads.
- Developed and validated ROS-based autonomous UCAV software through SITL simulation environments on Edge AI companion computers, while conducting extensive research and implementation work on both Single Object Tracking (SOT) and Multiple Object Tracking (MOT) systems for Vision-Based Autonomous UCAV operations.
- Developed an automated dataset generation and annotation pipeline leveraging the open-vocabulary segmentation model Segment Anything (SAM), significantly reducing manual labeling effort for huge datasets and accelerating YOLO training dataset preparation.
Computer Vision Engineer at North Tech (2025-10 – 2026-01)
Selected to contribute to large-scale 3D reconstruction and digital twin development projects based on a proposed computer vision solution for Ankara Foreign Affairs building project.
- Contributed to a computer vision-based construction monitoring platform that analyzed reconstructed site models against reference building models to assess construction progress and measure completion status with reduced manual oversight.
- Developed and optimized photogrammetry workflows using COLMAP while evaluating deep learning-based depth estimation approaches such as Depth Anything (DA3) for enhanced reconstruction performance and quality.
- Built high-resolution 3D models from datasets containing over 100,000 images using multi-GPU server environments (4× NVIDIA RTX 5090 or 4x Nvidia H100).
- Produced large-scale photogrammetric reconstructions and model-based progress analyses for a government infrastructure project, enabling digital monitoring of construction activities.
Software Development For UCAV Systems — AI, CV, Embedded Systems at Microntech Systems (2023-08 – 2025-06)
A defence company located in Istanbul, mostly working on UCAVs.
- Designed and implemented sensor fusion (e.g. KF, EKF, Mahony) algorithms, control algorithms for stabilization and autonomy on UCAV platforms.
- Developed GPU-accelerated deep learning systems for real-time object detection and tracking.
- Implemented and tested HITL simulations for testing developed software applications.
- Developed QT-based GUI Applications for manage embedded-to-host interfaces.
- Developed drivers for modules with communication interfaces like I2C, CANBUS, SPI, and UART
Remote Software Support Specialist at Sisal (2020-07 – 2020-08)
- Provided technical assistance and troubleshooting via remote tools.
