.png){kind=logo}
Artificial intelligence is transforming industries, shifting focus from traditional AI models to autonomous AI agents that make decisions and handle complex tasks. Businesses are leveraging these agents for customer support, supply chain optimization, and data analysis in various sectors, including healthcare.
However, as AI agents become more autonomous and complex, challenges in observability and control are now more essential than ever.
In the latest episode of the Analytics Insight Podcast, host Priya Dialani speaks with Datadog Field CTO of Asia Pacific and Japan, Yadi Narayana, about how observability will become an important aspect for organizations deploying AI-driven operations.
Yadi began by describing the company's vision, highlighting Datadog as the global leader in observability and security. It empowers over 30,500 customers worldwide, including 62% of Fortune 100 companies, to monitor and secure their digital ecosystems. From user experience and application performance to cloud infrastructure and security, Datadog offers a unified platform that enables businesses to operate efficiently.
In Q2 2025, Datadog announced a 28% year-over-year revenue growth to US$827 million, marking strong momentum. According to Narayana, Datadog's success is underpinned by ‘engineering-led innovation,’ with three times the R&D investment compared to competitors.
He pointed out that India is now a growth market with Datadog's operations increasing from a single employee to a team of over 100 professionals in Bengaluru, serving more than 300 customers and partnering with more than 30 partners.
"Observability has moved beyond mere server monitoring. It is all about having preventive, detective, and corrective visibility that supports large-scale digital ecosystems, especially in dynamic markets such as India, “the Field CTO of Asia Pacific and Japan in Datadog explained.
Asked to describe his role, Yadi said he was a "connector" between customers and the various R&D teams at Datadog. He's charged with ensuring observability and security deliver real business outcomes, not just technical metrics.
"The field CTO team co-creates multi-year observability strategies with enterprises," he said. "We help them accelerate innovation, build resilience, optimize costs, and enhance customer experience."
Datadog is collaborating closely with digital-native sectors in India, including FinTech, e-commerce, and BFSI, which emphasize reliability and scalability. Narayana also spoke about the community initiatives the company participates in to help bring together engineers, practitioners, and technology leaders.
Unlike static systems, AI agents can adapt their behavior based on context, input, and even random factors. So, observability performs a two-fold function:
Operational Monitoring: It helps detect performance bottlenecks, failures, or inefficiencies in real-time.
Continuous Improvement: The process of using telemetry data to fine-tune and retrain AI models to achieve superior results.
Traditionally, observability was seen as a back-end activity, ‘keeping the lights on.’ However, in an AI-first world, that perspective is no longer valid, Narayana highlighted.
"Every digital experience-from UPI transactions to video streaming-relies on speed and reliability," he said. "When observability prevents downtime or predicts anomalies, that's not just a technical success-it's a business win. Happier customers mean better retention and higher revenue."
Yadi further said that observability forms the data backbone for AI systems. Clean, contextual telemetry data is what makes AI models produce precise insights and not amplify noise. With Datadog's all-in-one platform, which includes logs, metrics, traces, and user signals, organizations unlock their full potential in driving AI-enabled operations.
While talking about the value of AI agent observability, Narayana shared a powerful example:
"In traditional setups, engineers might spend hours investigating outages by manually analyzing dashboards and alerts. But with AI-powered observability, the system itself can process billions of data points, pinpoint the root cause, and even suggest a fix within minutes."
He added that autonomous AI agents are now working like virtual teammates—automating routine investigations, cost optimization, and compliance checks. This not only enhances efficiency but also frees human teams to strive for innovation and strategic initiatives.
"Observability is no longer a side function; it's the foundation that makes AI effective," Yadi concluded. Organizations that treat observability as a business enabler unlock measurable productivity and long-term growth.
As AI agents are adopted into more enterprise workflows, the definition of success will largely be driven by observability. With organizations like Datadog paving the way for enterprise-scale observability, they can now apply AI systems that are transparent, accountable, and trusted. As Yadi Narayana explains, the observation of AI systems isn't just about the technology; it is about enabling business transformation.
