Building Conversational Genomics: Multi-Agent AI for Variant Interpretation
About this event
What if researchers could explore 7.8 million genomic variants through conversation instead of scripting?
This session dives into how we combined Google’s Agent Development Kit, Gemini, and Cloud infrastructure to transform variant interpretation from a code-heavy bottleneck into a real-time conversational experience. We’ll walk through the architecture, a spike-in validation test, and live demo with a codelab so you can start building with these patterns for your own research workloads.
Who Should Attend?
- Bioinformaticians & Researchers tired of context-switching between biological questions and Python scripts.
- AI Engineers & Developers looking for a blueprint to build reliable, multi-agent systems using Google’s Agent Development Kit (ADK).
- Workflow Architects interested in orchestrating high-performance scientific workflows.
Why Attend?
- See the Blueprint: Learn a reusable architecture for “Conversational Science” that works for genomics, climate data, or any heavy-dataset domain.
- Validate the AI: Discover how we designed a “spike-in” test to prove our agents could deliver clinical-grade precision without hallucinations.
- Build It Yourself: Get access to the codelab, with free cloud credits, and code patterns to start building your own research agents immediately.
