Agentic workloads put new pressure on both the data and observability layers. Databases must accommodate rapidly changing workloads without manual sharding or downtime, while engineering teams need visibility into model latency, token costs, GPU utilization, and end-to-end agent behavior.

In this on-demand technical webinar, AWS, TiDB, and Grafana Labs explore the architecture needed to build, monitor, and scale agentic applications in production, including technical walkthroughs and demos from both partners.

What you’ll learn:

• How distributed database architecture scales reads and writes without manual sharding
• How to support transactional, analytical, and vector workloads as agent activity grows
• How OpenTelemetry GenAI conventions help instrument agents without locking into a single AI stack
• How to correlate model performance, cost, agent behavior, infrastructure, and application signals
• How to monitor GPU and AI infrastructure utilization and cost at production scale

Featured speakers:

• Satish Subramanian, Manager, Partner Solution Architect, AWS
• Terry Purcell, Field CTO, TiDB
• Stephen Thorn, Head of Solution Engineering, North America, TiDB
• Shawn Pitts, Staff Technical Marketing Manager, Grafana Labs