I am a Computer Engineering student passionate about Machine Learning, Natural Language Processing, Large Language Models and Optimization Algorithms.
I enjoy creating projects that solve real-world problems.
I am currently focused on large language models (LLMs) for automated unit test generation and exploring the use of meta-heuristic optimization algorithms in areas such as prompt optimization and RAG (Retrieval-Augmented Generation) optimization.
In general, I investigate how LLMs can generate task-specific optimization algorithms to solve AI a problems efficiently.
| Languages | Frameworks & Libraries | Tools & Platforms |
|---|---|---|
| Python | PyTorch, TensorFlow, Hugging Face, LangChain, LangGraph, LangSmith, Chainlit, FastAPI, playwright, selenium | Git, GitHub, Docker, VS Code, Linux, Jupyter Notebook |
| — | — | Pinecone, Weaviate, Milvus, Qdrant , Chroma |
- Hyperparameter Optimization for CNNs – Implementing AI-based optimization algorithms for neural networks.
- DeepArticle – An AI-powered research and learning platform that discovers, analyzes, and summarizes information from academic papers, YouTube videos, and web sources. The system aggregates knowledge from multiple resources, provides intelligent search capabilities, and generates structured insights to help users learn complex topics faster and more effectively.
- RAGART – A customizable, production-grade RAG platform inspired by NotebookLM. It enables users to build AI-powered knowledge systems with configurable retrieval strategies, database backends, context management, and prompt engineering techniques. The platform supports advanced document processing, retrieval optimization, and multi-layer evaluation to improve answer quality and reliability.
- LinkedIn: linkedin.com/in/kadiryonak
- Email: kadiryonak7@gmail.com
- WebSite: kadiryonak.github.io
I enjoy reading books on philosophy and psychology, music production, and camping in my free time.