Future-Proofing AI Systems: Unlocking Business Value with Agentic AI Observability

The dawn of autonomous software is nigh - a revolution excitedly discussed at Transform 2025. Leading the dialogue were visionaries Ashan Willy, New Relic’s CEO, and Sam Witteveen, Red Dragon AI’s CEO and co-founder. The conversation focused on the sizable task of measuring and maximizing return on investment in agentic AI-systems. New Relic, in particular, provides invaluable ‘observability’ by capturing and correlating telemetry from applications, logs, and infrastructure in real-time.

Observability extends beyond mere monitoring - it’s about empowering teams with the significant context and insight required to understand, troubleshoot, and optimize intricate systems amid an unpredictable environment. With the rise of generative and agentic AI, this has become an increasingly challenging task. New Relic’s observability model now extends to monitoring entities like Nvidia NIM, DeepSeek, and ChatGPT, reflecting an accelerated AI deployment rate, currently at a 30% quarterly increase.

Observability in today’s era involves addressing unique use-cases across industries and ensuring functionality customized to each firm’s goals and size. For instance, a finance firm might concentrate on maximizing EBITDA margins, while a product-oriented company would focus on measuring quality control and market speed.

Since New Relic’s founding in 2008, AI has thrust observability back to applications with the emergence of agents, micro-agents, and nano-agents producing AI-written code.

As service and microservice numbers grow, especially for digital natives, the cognitive load for humans handling observability tasks is overwhelming. AI, Willy suggests, can alleviate this workload, working co-operatively, yielding critical information and easing the workload. The potential of agents in observability, he explains, is to automate workloads and democratize observability tasks, building it up to more people.

A unified platform for observability suits the agentic world perfectly. While agents simplify workflows, they create deep integrations across the complete ecosystem, including all tools that an organization uses, like Harness, GitHub, and ServiceNow.