UAlbany AI Lab Partners with New York Prosecutors Training Institute

Abdullah Canbaz and Mikail Demir stare at a computer screen inside the AI in Complex Systems Lab.
Abdullah Canbaz (left) and Mikail Demir review project files from inside UAlbany’s AI in Complex Systems Lab. (Photo by Zach Durocher)

By Mike Nolan

ALBANY, N.Y. (June 3, 2025) — Before Mikail Demir was a graduate research assistant at the University at Albany’s AI in Complex Systems Laboratory, he was a practicing lawyer in Turkey, where he studied law and economics.

Now, the first-year information science PhD student is combining his areas of expertise through a new partnership with the New York Prosecutors Training Institute (NYPTI).

NYPTI is a not-for-profit corporation created in 1995 by the District Attorneys Association of the State of New York. Its mission is to provide continuing legal education and mutual assistance to New York State prosecutors.

The new partnership between NYPTI and UAlbany’s AI in Complex Systems Lab, a cutting-edge research facility located on the second floor of ETEC, will focus on creating AI-driven tools that benefit the legal field.

Demir is working as a fellow through the partnership for the next year and a half.

“NYPTI is interested in using digital tools to streamline the legal research process for prosecutors and came to our lab to brainstorm ideas,” said Demir. “After meeting with us, they decided to hire a PhD student to work with them. With my background in law and artificial intelligence, it was a perfect project for me.” 

Mikail Demir points at a whiteboard inside the AI in Complex Systems Lab.
Mikail Demir is developing advanced large language models that can assist legal professionals. (Photo by Zach Durocher)

UAlbany’s AI in Complex Systems Laboratory was launched in spring 2023 by M. Abdullah Canbaz, an assistant professor in the Department of Information Sciences and Technology at UAlbany’s College of Emergency Preparedness, Homeland Security and Cybersecurity.

The lab is dedicated to the study and advancement of AI within the context of complex systems and networks. Demir is among around a dozen undergraduate and graduate students who conduct research at the lab.

Through the partnership, Demir will focus on two projects. First is an AI-driven decision summarizer tool that can generate accurate, concise summaries of criminal case decisions from NYPTI’s database. The tool, which functions similarly to ChatGPT, will rely on the use of advanced large language models (LLMs) to share the summaries. LLMs are trained on vast amounts of text data to understand and generate human-like language.

The second project is a legal conversational assistant, which will build on the decision summarizer, creating an interactive assistant that allows users to ask legal questions and receive case summaries with citations from the NYPTI databases. The tool will provide links to full case details and include a feedback mechanism for continuous improvement.

Demir will lead the collaboration on both projects with input from NYPTI and support from Canbaz, along with other student researchers at his lab.

“Before our initial meeting with Abdullah, we had started to explore AI and how it can help us with our busy caseloads. However, most of the commercially available tools were not easy to use and expensive,” said NYPTI Deputy Director Sean Smith. “This partnership was the missing piece of the puzzle for us. If successful, we believe it could be of great help to prosecutors across New York.”

“At the AI lab, we are interested in working with different industries and seeing how artificial intelligence can be integrated into their workflow in a meaningful and ethical way,” added Canbaz. “We are thrilled that NYPTI is working with us and funding a graduate student. It’s a great community partnership and shows our students are ready to work on real-world projects.”

Along with the NYPTI project, Demir also recently published a paper, along with Canbaz and UAlbany information science graduate student Hakan Otal,  that introduces LegalGuardian, a privacy-preserving framework for secure integration of large language models in legal practice. 

Demir says he plans to continue his work in artificial intelligence after graduation. 

“I think with my niche background in law, economics and information science that there will be opportunities for me,” Demir said. “I’d like to stay in the United States and work as a consultant who helps organizations harness the power of AI. It’s an exciting time to be in this industry.”