Speakers Agenda Why attend Partners Venue

Co-located ▼

Locations ▼

Get invited

Partnership opportunities

Call to action

Your text goes here. Insert your content, thoughts, or information in this space.

Button

Back to speakers

Vijay

Sanikal

Product Owner, Vehicle Simulation Integration

General Motors

Vijay Sanikal is a Product Owner at General Motors, with 15+ years of experience spanning automotive digital product development, CAE, cloud computing, synthetic data, AI, and thermal systems. His work sits at the intersection of Battery Electric Vehicles (BEVs) and Software-Defined Vehicles (SDVs), where he leads initiatives that reduce time-to-market, improve energy efficiency, and optimize vehicle performance using Software-in-the-Loop (SIL) models and AI-driven predictive analytics. Vijay holds a Master’s in Automobile Engineering from Anna University (MIT Campus) and an MBA in Marketing from Indiana University Kelley School of Business, giving him a rare blend of deep technical expertise and commercial perspective. His current focus is applying synthetic data and machine learning to BEV thermal systems—bridging research and real-world deployment while supporting global sustainability goals.

Button

15 April 2025 14:00 - 14:30

Tracing decisions across prompts, data, and tools

AI bugs are hard to reproduce because behavior is shaped by more than just input. Prompt state, retrieved context, tool availability, and execution order all influence outcomes, often in ways that aren’t visible once a request has completed. This session focuses on how teams trace and diagnose AI behavior in production. We’ll discuss how engineers follow decision paths across prompts, data retrieval, and tool calls, reconstruct execution context after the fact, and isolate the source of unexpected behavior. The emphasis is on practical techniques for debugging systems where behavior is probabilistic, stateful, and dependent on runtime conditions rather than deterministic code paths.