Prerna Kakkar is TL for Agentic Evaluation for Google Cloud Databases and is an active contributer to Evalbench and MCP Toolbox for Databases.
Using publicly available information we constructed an analysis to help you get a feel for this speaker before deciding to attend their session.
Kakkar is TL for Agentic Evaluation at Google Cloud Databases and a hands-on contributor to EvalBench and the open-source MCP Toolbox for Databases. Her AIE World's Fair 2026 session, co-presented with Kurtis Van Gent, offers a practitioner's framework for the build-time vs. run-time distinction in MCP tooling: why atomic developer tools become liabilities in production agents, and how to shift toward composite workflows, zero-trust workload identity, and agent-readable errors.