Ahmet Burak Kara

Engineer & Designer

About

As an avid engineering enthusiast, I'm constantly exploring and expanding my knowledge and skills in the field.

My passion for finding innovative solutions to complex problems has led me to work on exciting projects in electronics, mechanics, software development, and AI.

I'm particularly fascinated by the potential of AI to transform the way we live and work.

For me, engineering is not just a career, it's a passion.

Download my CV

Experiences

TUSAS Engine Industries Inc. (TEI) – Eskisehir, Turkey

Engineering Intern

August 2024 - September 2024

Worked with both embedded software design and hardware design teams.
Gained experience in developing safety-critcal aviaton software with DO-178C standards.

BMC Otomotiv Sanayi ve Ticaret A.Ş. – Izmir, Turkey

Engineering Intern

August 2023 - September 2023

Done my summer internshp at BMC Automotive's R&D department, involved in the development of an autonomous miltary vehicle.
Contributed to both hardware and software aspects of the project, collaborating closely wth interdisciplinary teams.
Gained experience using path-finding algorithms such as A* and RRT using the C/C++ programming language.

Projects

Distributed Computation on GPUs | TUSAŞ LiftUp | Supported by TÜBİTAK 2209-B | 2025

Developed a distributed training infrastructure enabling heterogeneous GPU nodes at TUSAŞ to collaboratively train deep learning models over a custom TCP-based communication layer. Reverse-engineered the YOLOv8 training pipeline to identify synchronization points and designed a socket-based protocol for efficient parameter and loss exchange across nodes. Implemented CPU/GPU benchmarking modules, hardware-eligibility checks, dynamic workload scheduling, and GPU-accelerated parameter aggregation to optimize performance on asymmetric systems. Experiments demonstrated up to 73% convergence toward ideal training time, significantly reducing model training duration and improving resource utilization. The project was funded by the TÜBİTAK 2209-B Research Support Program, recognizing its innovation and applicability within TUSAŞ’s distributed computing environment.

C/C++, Python, YoloV8

École 42 Piscine & Core Program

Selected for the École 42 Istanbul intensive Piscine program and successfully completed the highly competitive 1-month training based on peer-to-peer and project-driven learning. Expanded C programming proficiency by implementing core standard library functions from scratch and completing collaborative system-level tasks.After ranking high enough to pass the Piscine, advanced to the main curriculum, where I developed low-level memory manipulation algorithms and strengthened my foundations in problem-solving, debugging, and team-based software development.

C/C++, Shell, Algorithms and Data Structures

Autonomous Unmanned Surface Vehicle (Teknofest 2019 – Finalist)

During high school, completed electronics, Arduino, C, and Python training under the instruction of Dr. Tankut Akgül. Formed a three-person engineering team and competed in Teknofest 2019’s “Robotik Fetih 1453” challenge with an autonomous unmanned surface vehicle (USV).As team leader, coordinated project direction and team motivation while focusing on computer vision and systems integration. Developed the vessel’s object-detection and mission-execution capabilities using OpenCV, NumPy, and custom C-based modules. The USV successfully navigated between red vessels, autonomously retrieved designated green targets, and operated an onboard projectile mechanism to complete mission objectives. The project advanced to the finalist stage of the national competition.

C/C++, Python

Skills

Technical Skills

Control Systems Design
Robotics
ML/DL
Autonomous Systems

Software & Tools

C/C++
Python
SolidWorks
Proteus
AutoCAD
Adobe Photoshop

Languages

Turkish (Native)
English (Fluent)

Contact

akara.engineer@gmail.com linkedin.com/in/abkara github.com/ahmetbrkka

Writings

The Power of Minimal Design

Exploring how restraint and simplicity in design can lead to more powerful and memorable user experiences. Less is often more when every element serves a clear purpose.

January 15, 2025

Notes from Tokyo

Observations on design, culture, and the balance between tradition and modernity in one of the world's most dynamic cities. The attention to detail is remarkable.

January 10, 2025

Building Design Systems That Scale

Practical insights on creating and maintaining design systems that grow with your product while keeping consistency and clarity at the core.

December 28, 2024

On Creative Process

Personal reflections on developing a sustainable creative practice, managing inspiration, and finding clarity in the work that matters most.

December 20, 2024

Wandering Through Copenhagen

A week spent exploring Scandinavian design principles in their natural habitat, from architecture to everyday objects and public spaces.

December 5, 2024

Gallery

All Projects

A comprehensive collection of my work spanning engineering, design, development, and research initiatives.

Distributed Computation on GPUs | TUSAŞ LiftUp | Supported by TÜBİTAK 2209-B | 2025

Developed a distributed training infrastructure enabling heterogeneous GPU nodes at TUSAŞ to collaboratively train deep learning models over a custom TCP-based communication layer. Reverse-engineered the YOLOv8 training pipeline to identify synchronization points and designed a socket-based protocol for efficient parameter and loss exchange across nodes. Implemented CPU/GPU benchmarking modules, hardware-eligibility checks, dynamic workload scheduling, and GPU-accelerated parameter aggregation to optimize performance on asymmetric systems. Experiments demonstrated up to 73% convergence toward ideal training time, significantly reducing model training duration and improving resource utilization. The project was funded by the TÜBİTAK 2209-B Research Support Program, recognizing its innovation and applicability within TUSAŞ’s distributed computing environment.

C/C++, Python, YoloV8

École 42 Piscine & Core Program

Selected for the École 42 Istanbul intensive Piscine program and successfully completed the highly competitive 1-month training based on peer-to-peer and project-driven learning. Expanded C programming proficiency by implementing core standard library functions from scratch and completing collaborative system-level tasks.After ranking high enough to pass the Piscine, advanced to the main curriculum, where I developed low-level memory manipulation algorithms and strengthened my foundations in problem-solving, debugging, and team-based software development.

C/C++, Python, Algorithms and Data Structures

Autonomous Unmanned Surface Vehicle (Teknofest 2019 – Finalist)

During high school, completed electronics, Arduino, C, and Python training under the instruction of Dr. Tankut Akgül. Formed a three-person engineering team and competed in Teknofest 2019’s “Robotik Fetih 1453” challenge with an autonomous unmanned surface vehicle (USV).As team leader, coordinated project direction and team motivation while focusing on computer vision and systems integration. Developed the vessel’s object-detection and mission-execution capabilities using OpenCV, NumPy, and custom C-based modules. The USV successfully navigated between red vessels, autonomously retrieved designated green targets, and operated an onboard projectile mechanism to complete mission objectives. The project advanced to the finalist stage of the national competition.

C/C++, Python

Post Title

Date

Post content will be loaded here...