AI Con USA 2026 - MLOps & AIOps

AI agents have evolved beyond LLM chatbots; they possess the ability to plan, reason, and act autonomously. However, as their autonomy increases, understanding how they make decisions becomes more challenging. Traditional methods of observability—such as metrics, logs, and traces—capture outcomes but do not reveal the underlying reasoning. This session will explore how AI Agent Observability can shed light on the decision-making process by collecting and analyzing agent traces. We will discuss emerging standards like the Model Context Protocol (MCP), which provides structured and shareable...

Shipping AI into regulated, safety-critical mobile apps is more than “just call an LLM.” In automotive and medical contexts, we must meet strict privacy, latency, and auditability requirements while working with flaky connectivity and device constraints. In this talk, Ronak will share a field-tested architecture for agentic AI on the edge: a hybrid edge–cloud agent that uses tool-calling plugins, a policy engine for pre/post-prompt guardrails, offline fallbacks, and structured telemetry for audits. He will walk through how his team bounds agent behavior with capability cards, redacts PHI/...

As AI models grow in complexity and scale, inference efficiency has emerged as a critical engineering challenge for enterprise deployment. Traditional infrastructure built for training workloads often fails to meet the latency, throughput, and cost demands of large-scale inference operations. In this session, Sandeep will be sharing practical insights from engineering AI infrastructure at Broadcom, focusing on the end-to-end optimization of compute, networking, and storage subsystems. The talk explores techniques such as dynamic workload placement, adaptive batching, model quantization,...

Modern AI systems fail not only because of flawed models but because evaluation is often treated as a one-time task rather than an ongoing discipline. This session addresses the challenge of scaling evaluation across teams and pipelines to ensure model reliability, fairness, and performance. Drawing from real-world experience in large-scale financial analytics, Anusha Dwivedula will examine how product and data teams can collaborate to design a continuous evaluation framework that integrates precision, recall, and drift metrics with observability, lineage, and quality controls. She will...