In this AI Horizons episode, Rao Surapaneni, VP and GM of Google Cloud’s business application platform, joins Kathy Pham, VP of artificial intelligence at Workday, to discuss the ongoing evolution of enterprise AI and how leaders can turn it into measurable business value.
Emma Coppin
Multimedia Content Lead, Thought Leadership and Customer Stories
Workday
Audio also available on Apple Podcasts and Spotify.
The AI revolution is accelerating. Organizations across industries are weaving AI into products and operations while adapting to the speed of innovation. Some are already seeing breakthroughs, while many more are poised to capture new opportunities as their AI strategies mature.
In a recent episode of AI Horizons, Rao Surapaneni, VP and GM of Google Cloud’s business application platform, joined Workday VP of Artificial Intelligence Kathy Pham to explore what’s driving this moment.
Their conversation traces how decades of enterprise AI set the stage for today’s rise of interoperable agent systems and offers practical advice for leaders to deploy AI with confidence and impact.
Report
Surapaneni began his career at Bell Labs, working on early expert systems designed to predict faults in telecommunications networks—an early glimpse of predictive AI in action.
Over the decades since, he has seen wave after wave of AI innovation: expert systems, predictive models, generative AI, and now agents. Each sparked excitement, but also cycles of hype and disillusionment.
For Surapaneni, the difference between success and failure with new technology—whether it’s building the right products or deploying them effectively in the enterprise—comes down to alignment. Products and systems must be intentionally designed and implemented with a clear understanding of what value looks like and how it will be measured.
That focus, he notes, has guided his career.
“It’s always been fun to learn new technologies,” Surapaneni said. “Ultimately, building something useful and delightful to our customers has been what’s driven me. Throughout my career I’ve tried to place myself in situations where technology delivers value for customers in a full-blown product they can use.”
In other words: Technology alone isn’t the lever for transformation. As enterprises continue to adopt and integrate AI, success will depend on keeping strategy alignment and customer value at the center.
“To build an amazing product, you have to have your team working in cohesion across the full stack.” —Rao Surapaneni, VP and GM of Google Cloud’s business application platform
Building AI products requires rethinking what “full stack” means. In the past, it was shorthand for backend, storage, middleware, and front-end development. Today, the definition has expanded to account for data quality and provenance, the selection and tuning of large language models (LLMs), and orchestration of engineering, design, and product expertise.
Surapaneni emphasizes that building effective AI is a team effort. Technologists, scientists, data specialists, user experience (UX) designers, and evaluators must work in sync to ensure models perform as intended and interactions are intuitive. Just as importantly, product managers need the insight to balance trade-offs across business priorities, compliance, and customer experience.
“To build an amazing product,” Surapaneni emphasized “You have to have that team working in cohesion across the full stack.”
In his view, companies that treat the stack narrowly—as only technical infrastructure—miss the chance to deliver truly transformative AI. The real full stack is end-to-end: from raw data pipelines through thoughtful UX design to the continuous feedback loops that keep products improving over time.
Traditional software testing is deterministic: given the same inputs, you expect the same outputs every time. AI evaluation is probabilistic, and results shift depending on the data and context. Surapaneni explained that this requires a different mindset: one that’s rooted in experimentation, observation, and iteration.
Leaders must embrace practices like:
Together, these practices highlight that building trust in AI is as much about experience design as it is about model performance. Kathy Pham called out this point as one of the most important in Surapaneni’s process, noting UX as key for keeping organizations moving forward with intelligence.
“Never underestimate the power of user experience to help us move forward and understand our problems.” —Kathy Pham, VP of artificial intelligence, Workday
Google Cloud’s Agent2gent (A2A) interoperability protocol was created to solve a pressing enterprise AI challenge: enabling agents built on different platforms to communicate. By standardizing agent interoperability, A2A helps organizations maximize the full potential of agentic AI by enabling agents across platforms to complex complex tasks.
Today, the A2A protocol is managed by the Linux Foundation and governed as an open-source project where anyone can contribute. Adoption is accelerating. Major technology players like Cisco, IBM, and Wells Fargo have partnered together, signaling growing confidence in A2A’s role as the foundation for interoperable AI. The collaboration among competing standards and leaders highlights the growing understanding among technologists of the need to unite, ensuring customers and users benefit from one unified protocol and technology capability.
In the future, Surapaneni envisions an “internet of agents,” where interoperable systems don’t just exist inside a single company but collaborate across industries and even consumer contexts. He explained that the early focus has been intra-enterprise, where identity and trust are simpler because everything operates under one sign-on.
The next frontier is B2B integration, where supply chain, finance, and partner ecosystems must exchange information securely across company boundaries. Beyond that lies B2C and consumer-to-consumer scenarios, where the stakes for privacy and identity management grow even higher.
Achieving it will require strong foundations in trust, authentication, and security, but Surapaneni sees it not only as possible, but happening soon.
“We need to come together as technologists into one community in support of one A2A protocol so that customers can benefit.”
“Think big, start small, scale fast.” —Rao Surapaneni, VP and GM of Google Cloud’s business application platform
Surapaneni closed with guidance for leaders who want to seize AI’s full potential while avoiding the common pitfalls holding companies back. He emphasized that strategy and culture matter as much as technology. Going forward, companies must:
This combination of strategic clarity, workforce readiness, and iterative execution, he argued, is what will separate AI leaders from laggards in the years ahead.
A remarkable 82% of organizations are already using AI agents. But is your team ready? Read our latest report to learn how businesses are maximizing human potential with AI, featuring insights from nearly 3,000 global leaders.
Report
62% of leaders welcome AI, but only 55% of employees feel the same way. Closing this AI trust gap and ensuring that your organization is prepared for widespread AI implementation will prove decisive to your company's future success.