• Joined Leagle, a startup company founded by Ko¸c University students focusing on developing an AI-based legal assistant for businesses and lawyers.• Worked in implementation of assistant chatbot, contract drafting, and contract risk analysis features for the product.• Scraped data (legislations and regulations of Turkish Legal System) from public sources of government in order to fine-tune large language models and create database for linking model’s outputs to real-world sources.• Helped implementing this B2B service using Git Control System, Django framework, and Microsoft Azure with the team.

• Involved in a research group working on 360° panaromic image generation under the guidance of Prof. Aykut Erdem and Prof. Erkut Erdem.• Gained a deeper understanding of autoregressive multimodal models and stable diffusion models.• Research was focused on vision language models such as Anole and Chameleon.• Experimented with different tokenizers and image generation strategies.• Extended existing models with new techniques and performed fine-tuning.

• Joined in a research group in Leiden Institue of Advanced Computer Science (LIACS) under the guidance of Asst. Prof. Hazel Doughty.• Worked on motion recognition and video understanding with large multimodal models.• Tested various LLaVa models, including LLaVa-NeXT and Video-LLaVa, common vision language models built for image and video recognition. • Fine-tuned models in action recognition datasets such as Something-Something videodataset.• Utilized models like BERT-Sentence Transformer to evaluate the models.• Learned how to implement quantization and LORA for efficient and faster training.• Adapted training codebases for multi-GPU training.

• Completed a one month game development internship in Atlas Space, a company focusing on building technology in the fields of VR/AR, metaverse platforms, IAAS, and blockchain• Joined the Game Development and Interactive Media department and worked as a software/game developer intern.• Mastered my C++ and Unity skills.

• Involved in a research group of Asst. Prof. Fatma Güney working on on motion forecasting for self-driving cars.• Tested the robustness of prediction models such as QCNet and HPTR by manipulating traning and validation datasets.• Worked with Argoverse 2 and Waymo Open Dataset in a cluster system.• Reproduced machine learning papers and gained hands-on experience in implementing neural networks.• Mastered my Pandas, NumPy, matplotlib, and PyTorch skills.

• Worked as a teaching assistant tutor for Linear Algebra course offered by Prof. Burak Özbağcı in Fall 2023.• Provided one-on-one or group teaching and problem-solving sessions for students.