Data Generation Automation for Q&A LLM Models:I developed a local data production automation that speeds up the process of fine-tuning Q&A LLM models 21 times by extracting context, question and answer structures from documents. This system optimized the workflow and provided efficiency.Customer Representative Project with Human-like Gestures and Speech:Using the Llama-3 and Llama-3.1 models, I developed an agent with human-like gestures and natural speech capabilities. This project aims to provide natural and effective communication in customer interactions.Agent Based Systems:Using the Llama-3.1, Llama-3, and Mistral Large models, I prototyped agent-based systems that interoperate with the ReactAgent library to create micro LLM frameworks that can efficiently manage complex tasks and work in concert with each other.Fine-Tuning and Testing of Llama-3.1 and Llama-3 Models:I fine-tuned the Llama-3.1 and Llama-3 models and tested them locally. In this process, I performed detailed tests to optimize the performance of the models and make them suitable for application scenarios.Dual Authentication with Voice Bio-sequencing:I added a second layer of verification with voice bio-authentication in addition to password logins for mobile banking applications. This system provides a secure verification by integrating with solutions such as mobile signature.Synthetic Data Generation Automation for High Quality LLM Datasets:I added an extra similarity algorithm to the synthetic data generation automation, enabling the creation of unique and high quality LLM datasets. This increases the quality and diversity of datasets.Function Calling and Micro LLM Structures to Improve Missing Skills:I improved the weaknesses of systems with weak mathematical skills, such as Llama-3, to expert level with function calling methods. I strengthened the weaknesses of the models by creating micro LLM structures with agent-based systems.

In Ozteknik, I aim to provide effective and efficient contributions to the decision-making processes of the factory, to improve the factory and business processes by converting data into meaningful results.

By analyzing Businta's extensive data sets, I gain valuable insights into optimizing business processes and decision-making mechanisms. By turning data into understandable and interactive visuals, I make it easier for decision makers within the company to make data-based decisions. I am developing special machine learning models for the needs of companies. By developing artificial intelligence-based solutions, I help solve the company's current and potential business problems.

I am developing projects in the field of organizing, visualizing and modeling big data at Kapsül Teknoloji. I am also working on an artificial intelligence that can detect sinkholes that may occur in the future and I am conducting joint projects with the Information Processing Department. I believe that I will achieve many great successes at Kapsül Teknoloji. If I have to talk about the projects I work on at Kapsül, 360 virtual tour of Kapsül Teknoloji, moving the trainings to be held on Kapsül Teknoloji to virtual reality environment, Kapsül ChatBOT that can get information while interacting with people and a local robot production to ensure communication inside like JARVIS in Iron Man movie.

I am developing projects in the field of editing, visualization and modeling of big data in Kapsül Teknoloji.

I am working on artificial intelligence projects that can for health, industry, and businesses that can lead the future of technology.

Planning a building's electrical equipment and building electricity with AutoCad.