October 7, 2026 at 12:00 PM ET

Transforming Research Data Access with GenAI

A Practical Multi-Agent Text-to-SQL Approach

Research administrators, staff, and institutional leaders frequently need timely insights into data, in this case, institutional research data. This may include Proposal, Award, and Expenditure data, with questions ranging from understanding department- and PI-level collaboration and major research areas of interest to grant award trends and detailed expenditure patterns. While this presentation focuses on institutional research data, the approach can be applied to a wide range of data and use cases.

Traditional approaches such as manual analysis, custom reports, and static dashboards are effective for predefined metrics but can be limiting when stakeholders need answers to ad hoc questions. This presentation introduces a practical Generative AI (GenAI) solution that functions as an intelligent, ondemand research data analyst. The tool provides secure access to proposal, award, and expenditure datasets and uses a multi-agent architecture to route questions to specialized Proposal, Awards, and Expenditure Agents, coordinating across agents when needed. It dynamically generates and executes Python code to query the underlying data and deliver contextual answers to natural-language questions in real time, enabling faster decision-making and more flexible exploration of institutional data.

The session will include a live demonstration of the tool, along with practical implementation considerations, including the cloud resources and services used to build the solution, the cost of building it, and the technical skills required for a first-time implementation. The presentation will also address how data security, accuracy and validation safeguards, and responsible AI governance have been incorporated into the solution.

What You'll Learn

By the end of this session, participants will be able to:

1. Understand how a Generative AI (GenAI) multi-agent solution can transform access to institutional data by enabling natural-language queries on tabular datasets.
2. Explore the architecture and workflow of the solution, including specialized agents, query routing, multi-agent coordination, and dynamic Python code generation for accurate query responses.
3. Learn how to adapt and apply this GenAI framework to their own institutional use cases, data sources, and research administration needs.
4. Evaluate critical implementation considerations, including data security, accuracy and validation safeguards, cost, technical requirements, and responsible AI governance.

Prerequisites

Attendees are encouraged to have some familiarity with programming languages, technical concepts, and a basic understanding of Large Language Models (LLMs). That said, the session is open to IT professionals, research administrators, institutional leaders, and anyone curious about how GenAI can be applied to research data and other institutional use cases.

Meet the Speaker

Chandni Mathur is Data Analyst on the Research Analytics team at the University of Illinois. She works across SQL, Python, and Linux to provide analytical support, and builds and maintains the Tableau dashboards that track the key metrics behind the university's proposals, grants, and student data.

She will show how the agents work together and why splitting the job across agents makes the output more reliable than a single model guessing at a query. She'll cover where this fits in real research-data workflows, what it takes to trust the answers, and where a human still belongs in the loop.

Save Your Seat

Space is limited. Register now to join the conversation and get your questions answered live by the team behind the research.