Kadir Göksel
Gündüz.
AI/ML Engineer & Technical Lead
Istanbul, Turkey
Innovative AI/ML Engineer specializing in machine learning, deep learning, and natural language processing. I develop custom neural architectures and scalable frameworks for Computer Vision applications through cutting-edge technology. My expertise lies in optimizing feature selection and neural network components to deliver high-performance intelligent systems that solve real-world problems. I transform complex technical challenges into practical, impactful solutions that drive measurable business outcomes across various industries.
Experience
Professional journey in AI engineering, technical leadership, and research
Autonomous Farming Vehicles Venture Leader
Leading the development of autonomous agricultural vehicles under the TUBITAK BIGG program.
- Executing SLAM and ROS integration for autonomous farming vehicles as part of a personal venture project under the TUBITAK BIGG program
- Developing swarm intelligence algorithms alongside AI integrations for the vehicles
- Managing patent procedures for the prototypes being developed
AI Engineer & Team Lead
Led a team integrating LLM-powered AI into a SaaS-based educational platform with cloud-native architecture.
- Led team in integrating LLM into SaaS-based educational platform for personalized learning experiences
- Managed deployment pipeline on AWS infrastructure (EC2, S3, Lambda, SageMaker)
- Developed TrOCR-based handwriting recognition system for digitizing student submissions
- Implemented ORPO/DPO alignment training for fine-tuning domain-specific language models
- Built PostgreSQL semantic search engine for intelligent content retrieval and recommendations
Generative AI Engineer
Developed and deployed Virtual Try-On AI system using generative models and cloud infrastructure.
- Built Virtual Try-On AI system enabling realistic clothing visualization on user images
- Implemented LoRA fine-tuning techniques for efficient model adaptation and domain specialization
- Deployed inference pipeline on AWS EC2 with optimized latency for real-time processing
- Integrated MongoDB for managing user sessions, model metadata, and inference results
Machine Learning Engineer
Data analysis and AI development for optimizing wind turbine fleet operational efficiency.
- Developed energy forecasting models and executed performance optimization processes
- Designed comprehensive control dashboards with PowerBI for wind turbine monitoring and data analysis
- Built domain-specific LLM models for corporate needs and strategic objectives
- Optimized corporate information processing through custom-trained models on company data
- Developed automated document analysis and classification systems using NLP techniques
AI & R&D Engineer (Intern)
Spectrum analysis, image processing, deep learning model development, and real-time audio classification systems.
- Led project team in developing classifier models for gender and age classification using audio spectrums with time series, anomaly detection, and signal processing methods
- Developed TFGAN (Time-Frequency Domain) model for synthetic voice generation and voice cloning
- Built real-time audio classification pipeline for live signal processing
- Developed Speech-to-Text system components with custom language model integration
- Designed QT-based graphical user interface for audio analysis and visualization tools
Projects
Research areas and technical projects across AI domains
- Built and deployed Turkish paragraph-level QA model for production use
- Optimized inference latency for real-time response generation
- Commercially deployed and serving live user queries
- Implemented ORPO/DPO alignment training for domain-specific LLMs
- Evolutionary optimization for hyperparameter search
- Explored Mixture of Experts (MOE) architectures
- Domain-specific LLM models for corporate requirements
- Automated document analysis and classification using NLP
- Gender and age classification from audio spectrograms
- Speech-to-Text system with custom language model integration
github : @RsGoksel
linkedin: kadir-goksel
email : gokssel.gunduz@gmail.com
View Resume →
Full CV with experience, publications, and technical skills.
2from transformers import AutoModel
3
4# load pretrained weights
5model = AutoModel.from_pretrained(
6 "RsGoksel/custom-vision"
7)
8model.eval()
Open Source Projects →
AI/ML repositories, computer vision, reinforcement learning, and autonomous systems.
Project Timeline
A chronological view of research and development work
Virtual Try-On AI System
Apr – Aug 2025Educational LLM Platform with TrOCR
2024 – 2025PostgreSQL Semantic Search Engine
2024 – 2025TEKNOFEST AI Competition — 4th Place
2024AWS Cloud & Serverless Architecture
2024Turkish QA Model (Commercial)
2024Advanced LLM Conversation System
2024Wind Turbine Energy Forecasting
Dec 2023 – Mar 2024ORPO/DPO Model Alignment
2024UMAY Autonomous Farming Launch
Aug 2023CV for Autonomous Navigation
Dec 2023Transformer NLP for Medical Data
Sep 2023ESRGAN Super Resolution (MSc Thesis)
2023Forward-Forward Algorithm + PPO
2023Transformer Policy Optimization
2023TFGAN Voice Synthesis & Cloning
Jul – Sep 2022RL for Robotic Control (Unity A2C)
2022Image-to-Image GAN (Canik Municipality)
2022Swarm Intelligence Path Planning
Mar 2022Speech-to-Text & QT Interface
Jul – Sep 2022YOLO-Inspired Detection Framework
2021CNN-Transformer Hybrid Architecture
2021Time Series Neural Network
Oct 2021Writing
Technical articles and blog posts published on Medium
Fundamental Terms of Signal Processing
An exploration of essential concepts and terminology in signal processing, providing a solid foundation for engineers and researchers.
Sinyal Isleme Temel Kavramlari
Sinyal isleme alanindaki temel kavramlarin Turkce aciklamalari ve ornekler ile desteklenen kapsamli bir rehber.
Particle Swarm Optimization
A deep dive into Particle Swarm Optimization algorithms, exploring their applications in solving complex optimization problems.
SNN Aglari
Spiking Neural Networks (SNN) hakkinda detayli bir inceleme ve bu gelisen yapay zeka teknolojisinin uygulama alanlari.
Get in Touch
Feel free to reach out for collaborations, research opportunities, or just to connect