I'm a software developer focused on applied AI, LLM-powered products, computer vision, and edge intelligence.

My work sits at the intersection of AI engineering, product thinking, and real-world deployment. I like building systems that are not only technically interesting, but also usable, explainable, and connected to actual user needs.

• LLM Applications

  • RAG pipelines
  • Prompt engineering
  • LangChain-based workflows
  • Model evaluation and response quality analysis

• Machine Learning & Deep Learning

  • Classification pipelines
  • Transformer-based NLP experiments
  • Fine-tuning and benchmarking small language models
  • Model comparison, evaluation metrics, and reproducible experimentation

• Computer Vision & Edge AI

  • OpenCV-based pipelines
  • Image-derived feature analysis
  • Visual tracking and autonomous edge systems
  • AI modules for real-time field applications

• Product-Oriented AI

  • Turning AI capabilities into user-facing product features
  • Working across research, development, testing, and iteration
  • Building systems with both technical and user experience constraints in mind

I worked on Gipi, an AI companion app that reached 750K+ users across 170+ countries. My role involved product ownership, AI behavior design, prompt systems, RAG pipelines, and model evaluation for empathetic and context-aware conversations.

Featured on NVIDIA Developer Blog:
Personalized Learning with Gipi: NVIDIA TensorRT-LLM and AI Foundation Models

I have been working on applied AI systems involving computer vision, visual tracking, and real-time decision-making constraints. My current interests include edge AI, autonomous systems, and deploying lightweight intelligence closer to the field rather than keeping everything in cloud-only workflows.

My current academic work includes machine learning and computational intelligence projects such as:

• Date fruit multiclass classification using Artificial Neural Networks
• Fake news detection and reproduction studies with transformer-based models
• Turkish answer quality evaluation using small language models
• Dataset analysis, preprocessing, model training, and evaluation workflows

Languages: Python, Java, JavaScript, HTML/CSS
AI/ML: TensorFlow, Keras, Scikit-learn, PyTorch, Hugging Face, LangChain
Data: Pandas, NumPy, Matplotlib, Seaborn
Computer Vision: OpenCV
Backend: Flask, REST APIs
Tools: Git, GitHub, Jupyter Notebook, Google Colab, VS Code
Other: Unity, object-oriented programming, technical documentation

I'm currently focused on building stronger applied AI projects around:

• LLM evaluation
• Small language models
• Turkish NLP
• Computer vision
• Edge AI and autonomous systems
• Reproducible ML workflows

I care about communication, clarity, and collaboration. My background includes teaching, product ownership, and software leadership, which shaped how I approach technical problems: systems that people need to understand, use, and trust.

• LinkedIn
• Email

Building AI systems that are useful, understandable, and real-world ready.