Ahmet Doruk Gonulal (Ahmet Doruk Gönülal) is an entrepreneur, AI developer, and student researcher based in Izmir, Türkiye. He is the co-founder and CEO of Kedy Technologies, an AI research startup focused on efficient, domain-specific language models and intelligent systems. His work brings together company building, applied machine learning, and robotics.
His technical experience spans natural language processing, computer vision, reinforcement learning, and robot software. Across these areas, he focuses on a common question: how can intelligent systems turn language and sensory information into useful decisions and executable actions?
Biography
Doruk combines entrepreneurship with hands-on work in artificial intelligence and robotics. As co-founder and CEO of Kedy Technologies, he is building an AI research startup around efficient, domain-specific language models and intelligent systems. His entrepreneurial interests are closely tied to his technical work: identifying specialized problems where machine learning can support useful applications in engineering, data analysis, and autonomous systems.
He is currently completing his education at Izmir Bahcesehir Science and Technology High School. Alongside his formal studies, he has independently pursued machine learning, calculus, and advanced mechanics. These subjects connect the mathematical foundations of learning algorithms with the physical systems he works on in robotics.
His project experience spans natural language processing, computer vision, reinforcement learning, autonomous vehicles, and AI agents. Across these fields, he is particularly interested in the connections between components: how perception informs a decision, how a decision becomes an action, and how feedback from an environment can inform the next step. This systems-level interest underpins his focus on AI that can work beyond text generation.
As Electronics and Programming Senior Lead of FRC Team 6429 — 4th Dimension, Doruk works on robot software, perception, autonomous navigation, and control systems. His role brings technical leadership together with implementation across software and electronics, connecting the interpretation of sensor information to the behavior of a physical robot.
He represented Türkiye as a National Team Member at the 2026 European Olympiad in Artificial Intelligence (EUROAI), finishing 29th overall. His work in AI education also includes serving as Education Committee President at CicekliAI, where he contributes to workshops on machine learning foundations and healthcare data analysis. The workshops connect model development with data preparation, evaluation, and the limitations of applying AI to practical problems.
His current direction brings these experiences together: developing smaller, specialized models and agentic systems that connect language, perception, and action. He is interested in how these systems can become practical tools within technical workflows, linking his research interests to the applications he aims to build as a founder.
Entrepreneurship and experience
Co-founding Kedy Technologies
At Kedy Technologies, Doruk combines his role as co-founder and CEO with a technical focus on efficient AI and domain-specific language models. The startup connects his entrepreneurial work with his interest in building intelligent systems for specialized applications.
His focus is on the relationship between model capability and practical utility: how specialized systems can support work in engineering, data analysis, and other technical environments. This interest informs both his research direction and the kinds of applications he aims to build.
Technical leadership in robotics
As Electronics and Programming Senior Lead at FRC Team 6429, Doruk works across robot software, perception, autonomous navigation, and control. These areas connect the interpretation of sensor data with decisions about how a physical robot should move and respond to its environment.
| Organization | Role | Area |
|---|---|---|
| Kedy Technologies | Co-Founder & CEO | Entrepreneurship & AI research |
| EUROAI | National Team Member, Türkiye (2026) | Artificial intelligence |
| 4th Dimension · FRC 6429 | Electronics and Programming Senior Lead | Robotics |
| CicekliAI | Education Committee President | Education |
A full resume is available on request.
Workshops
As Education Committee President at CicekliAI, Doruk contributes to technical education on artificial intelligence in healthcare. His workshops connect the foundations of machine learning with the practical work of preparing datasets, developing models, and evaluating their results.
Artificial Intelligence in Health: Foundations
An introductory course covering the fundamental concepts of artificial intelligence and machine learning in healthcare. The course explores how AI systems analyze medical data, support clinical decision-making, and address healthcare-related problems. Topics include supervised and unsupervised learning, model development workflows, healthcare datasets, evaluation metrics, and ethical considerations.
Artificial Intelligence in Health: Data Analysis
A practical course focused on the analysis and preprocessing of healthcare data for machine learning. Topics include cleaning, organizing, visualizing, and interpreting medical datasets, statistical analysis, feature engineering, and handling missing or imbalanced data for predictive modeling.
See also: CicekliAI on YouTube.
Research
Ahmet Doruk Gonulal’s research interests connect efficient machine learning with autonomous systems. He is interested in the full path from language and perception to decision-making and action, particularly in specialized environments where a model’s output must support a concrete task.
Efficient and specialized language models
Through Kedy Technologies, Doruk focuses on small and domain-specific language models. His interests include the trade-offs between model size, computational requirements, and task-specific capability, and how specialization can make language models useful in engineering and data analysis workflows.
Agentic systems and reinforcement learning
Doruk is interested in systems that connect language-based reasoning with sequences of actions and feedback from an environment. His work and interests in AI agents and reinforcement learning center on how an intelligent system chooses what to do, observes the result, and uses that information in subsequent decisions.
Robotics and embodied AI
Doruk’s robotics work connects computer vision, autonomous navigation, and control systems. At FRC Team 6429 — 4th Dimension, he works on software that links perception to robot behavior. His broader interest in embodied AI extends this connection to language-guided systems that can act in physical environments.
For research discussions and collaboration, contact Ahmet Doruk Gonulal.
Contact
For startup collaboration, technical partnerships, or research discussions: [email protected].