Eser Kandogan
About Eser Kandogan
Eser Kandogan is a Principal Research Engineer known for his contributions to fairness-aware data preparation frameworks and multimodal data discovery in AI systems. He has co-authored several papers, including work on ExtremeReader and the sensitivity of language models.
Work at Megagon Labs
Eser Kandogan serves as a Principal Research Engineer at Megagon Labs. In this role, he focuses on advancing research and development in the field of artificial intelligence and data processing. His work includes the development of innovative frameworks and tools that enhance the capabilities of AI systems.
Research Contributions
Kandogan has made significant contributions to various research projects. He was involved in developing a fairness-aware data preparation framework for entity matching, which aims to improve the equity of data handling processes. Additionally, he worked on a collaborative annotation approach that integrates large language models with human verification to enhance data accuracy.
Publications and Presentations
Eser Kandogan co-authored several papers in the field of AI and data science. Notably, he contributed to a paper on ExtremeReader, which was presented at the WWW 2020 conference. This work focused on creating an interactive explorer for customizable and explainable review summarization. He also co-authored research on the sensitivity of large language models and reasoning capacity in multi-agent systems.
Benchmark Development
Kandogan played a key role in the creation of CMDBench, a benchmark designed for multimodal data discovery in compound AI systems. This benchmark aims to facilitate the evaluation and comparison of different AI models and systems, enhancing the understanding of their performance in complex environments.
Architectural Frameworks for AI Systems
He has contributed to the development of a blueprint architecture for compound AI systems tailored for enterprise settings. This architectural framework is intended to guide organizations in implementing AI solutions that are efficient and scalable, addressing the unique challenges faced in enterprise environments.