IEEE Graduate Student Body @ Ohio State
The IEEE Graduate Student Branch (GSB) at The Ohio State University is a graduate-led organization dedicated to fostering technical excellence, leadership, and professional development among graduate students.
IEEE OSU AI Day 2026
We are organizing the IEEE OSU AI Day on February 12, 2026, hosted by the
IEEE Graduate Student Body (GSB) and the IEEE Undergraduate Chapter.
Join us for talks and discussions featuring invited speakers across AI research and real-world applications.
🍽️ Free lunch and refreshments will be provided.
1739 N High St, Columbus, OH 43210
Program Timeline
IEEE GSB overview, event goals, speaker introductions
Networking with speakers and attendees
Speakers
We have five speakers lined up for the event. Below are their photos, talk abstracts, and short bios.
Balavignesh Vemparala
Title/Affiliation: Senior R&D Engineer, Synopsys (formerly Ansys)
Talk Title: From LLMs to Agentic AI: Building Reliable Systems with RAG and MCP
Abstract
Large Language Models (LLMs) are powerful, but the most useful applications come from the systems built around them. This talk explains the modern LLM stack that turns a model into something dependable and action-capable: Retrieval-Augmented Generation (RAG) to ground outputs in trusted information, and Agentic AI to plan and execute multi-step workflows with tools and verification. We’ll break down practical RAG design choices (chunking, retrieval, reranking, and evaluation), then walk through agent patterns used in industry—tool-using assistants, planner–executor setups, manager–worker teams, and evaluator/guardrail agents. Finally, we’ll introduce MCP (Model Context Protocol) as a structured way to connect models to tools and data sources, reducing one-off integrations and making agent systems easier to scale. Attendees will leave with clear mental models, real design patterns, and a roadmap for building portfolio-ready LLM applications.
Bio
Balavignesh Vemparala is a Senior R&D Engineer at Synopsys (formerly Ansys), where he develops high-performance computational methods for large-scale engineering simulation. His work focuses on core finite element solver development in Mechanical APDL, HPC performance enhancement, and integrating AI/LLM-assisted automation—such as retrieval-augmented generation (RAG) and agentic AI workflows—to streamline simulation setup, pre-/post-processing, and improve user productivity. He earned his Ph.D. in Mechanical Engineering from The Ohio State University, where his research spanned computational mechanics and scientific machine learning. His interests lie at the intersection of numerical methods, high-performance scientific computing, and AI for engineering applications.
Dinesh Kumar Garg
Title/Affiliation: Senior IT Manager, Honeywell Intelligrated
Talk Title: AI-Enabled Real-Time Inventory Reconciliation in Enterprise Systems
Abstract
Enterprise resource planning (ERP) and execution systems often fail to maintain accurate, real-time
alignment between system inventory records and physical operations. Inventory drift caused by fragmented
processes, delayed transactions, and manual interventions reduces trust in system data and limits the
effectiveness of downstream planning, optimization, and automation initiatives.
This talk presents an applied AI approach to real-time inventory reconciliation in enterprise systems,
combining unified process design with layered AI capabilities: anomaly detection to surface inventory
deviations, time-series and risk-based prediction to assess downstream impact, and constraint-aware
decision logic to recommend corrective actions. Retrieval-augmented generation (RAG) is used selectively
to support contextual knowledge retrieval and explanation, assisting users with decision context without
replacing deterministic operational models.
Bio
Dinesh Kumar Garg is an industry researcher and IT, supply-chain, and advanced manufacturing technology leader focused on applying artificial intelligence to enterprise resource planning and execution challenges. He is a Senior IT Manager at Honeywell Intelligrated, where he leads ERP, advanced planning, and digital manufacturing initiatives supporting complex industrial operations. His work centers on AI-driven inventory optimization, supply-chain planning, and real-time, data-driven production decision-making. He is a Senior Member of IEEE and an author and contributor to CIO.com. Garg’s background spans applied research and large-scale industrial system delivery across multiple industries. Earlier in his career, he was associated with the Central Scientific Instruments Organization (CSIO), a national laboratory under India’s Council of Scientific and Industrial Research (CSIR). He later contributed to enterprise ERP and manufacturing initiatives for General Electric (GE) through Birlasoft and supported advanced manufacturing and supply-chain planning initiatives at ADVICS North America, a Toyota–AISIN company. His work has resulted in multiple U.S. and UK patent filings related to AI-enabled inventory and supply-chain optimization.
Deepak Warrier
Title/Affiliation: JPMorgan Chase
Talk Title: AI at Scale: Navigating Innovation and Security at JP Morgan Chase
Abstract
This session explores the operationalization of artificial intelligence within JPMorgan Chase, navigating the shift from experimental research to large-scale industrial deployment. We will discuss real-world applications within cybersecurity, including NLP for automated threat classification and proactive threat modeling workflows. The talk will also cover constraints in the financial sector—least-privileged access, regulatory auditability, and cost-efficiency—and conclude with trends such as agentic workflows and scalable, explainable AI infrastructure.
Bio
Deepak has worked on operationalizing machine learning at scale, including NLP systems for compliance and production ML infrastructure. He is a co-author of research on AI-assisted threat modeling and MITRE ATT&CK tagging. He holds Bachelor’s and Master’s degrees in Computer Science from The Ohio State University and has previously held engineering roles at Salesforce.
Nitin Appiah
Title/Affiliation: Data Analyst Hexion
Talk Title: Applied Artificial Intelligence in Chemical Manufacturing
Abstract
Manufacturing industries are rapidly adopting AI and transforming into more technology-focused organizations. This talk presents a set of applied AI initiatives at Hexion, including time-series forecasting for demand planning using macroeconomic signals, machine learning–driven product formulation to tailor chemical characteristics, and the use of streaming manufacturing sensor data to create digital twins. The session also explores large language models and context-aware prompt engineering to automate reporting and improve insight generation from complex enterprise data. Attendees will gain practical perspectives on translating academic AI concepts into scalable, production-ready solutions within a real-world chemical manufacturing environment.
Bio
Nitin Appiah is a Data Analyst in the enterprise data analytics team at Hexion Inc., where he applies machine learning, forecasting, and generative AI techniques to large-scale industrial and business systems. His work includes LLM-driven analytics automation, predictive modeling, and the development of production-grade data pipelines and BI systems. He holds a Master’s degree in Computer Science from The Ohio State University, USA (2023), and a Bachelor’s degree in Computer Science from Anna University, India (2021), with formal training in machine learning, neural networks, and parallel computing. Previously, he interned at Synopsys in California and Vue.ai in India. He also served as a Graduate Research Assistant in Ning Lab at The Ohio State University, conducting healthcare AI research to improve treatment for patients with traumatic brain injury and Alzheimer’s disease. He has additionally contributed to research on federated learning, multilingual language models, and computer vision.
D. Adithya Sriram
Title/Affiliation: Engineer, Demand Side Operations, Dominion Energy
Talk Title : ML and probabilistic modeling techniques for Power Systems
Abstract
Rapid load growth across the utility industry has intensified the need for accurate short and long term forecasting methods that can predict future capacity and system planning requirements. This talk presents an exploration of how ML and probabilistic modeling techniques can be integrated into time series forecasting workflows to enhance capacity planning and support flexible load programs within modern power systems. This session will highlight how future transmission reliability can be inferred from current system operating conditions through models capable of forecasting N-1 security constraints. Additionally it will also touch upon how AI agents are being used to autonomously respond to response tasks where many questions regarding utility programs recur annually across different stakeholders.
Bio
D. Adithya Sriram is an Engineer within the Demand Side Operations team at Dominion Energy. He works on challenges at the intersection of power system reliability and grid operations, including forecasting and functional technology to maintain dependable performance across the transmission network. He holds an M.S. in Electrical and Computer Engineering with specialization in power systems, signal processing, and control systems, and is passionate about bridging academic research and utility-scale implementation.
Organizing Committee
Organized by the IEEE Graduate Student Body (GSB) at The Ohio State University.
