Google, doubling down on synthetic intelligence (AI) to gasoline income development, is witnessing elevated enterprise curiosity in agentic workflows, based on Oliver Parker, vice chairman of world Generative AI (GenAI) go-to-market at Google Cloud.

In the previous few months, the curiosity has been rising considerably, Parker famous throughout a latest interview in India.

AI brokers—autonomous methods able to decision-making and motion—are designed to carry out duties with minimal human intervention. From driverless automobiles to good residence assistants and buying and selling bots, these brokers characterize a important evolution in AI know-how.

Learn this | AI and GenAI are remodeling robotics: ABB robotics world head

Google’s newly launched AgentSpace platform permits companies to create AI-powered workflows that combine seamlessly into operations, providing instruments for duties like enterprise search and information interplay. It’s “taking AI past builders to the enterprise layer,” Parker defined.

Over the previous yr, Google has poured assets into growing agentic fashions. These methods can course of broader contextual information, anticipate a number of steps forward, and execute duties autonomously underneath human supervision. CEO Sundar Pichai underscored this shift in a December weblog submit saying Gemini 2.0, describing it as the muse for the “subsequent period of fashions constructed for this agentic period.”

Constructing agentic methods in low-code environments is rising because the quickest technique to drive enterprise worth.

When to make use of AI brokers?

The worth of agentic methods lies of their skill to commerce off latency and price for enhanced process efficiency, and people utilizing these methods ought to think about when this tradeoff is smart, famous a 20 December weblog by tech firm Anthropic.

Nevertheless, firms must deploy these methods judiciously, Parker stated. “Consider a mannequin as an engine and an agent as a automotive,” he defined, emphasizing that many enterprises embed massive language fashions (LLMs) into workflows with out creating absolutely autonomous methods.

For cost-sensitive and latency-critical use instances, Google provides Gemini Flash, a mannequin optimized for pace and affordability. In distinction, Gemini Professional caters to reasoning-intensive duties, albeit at increased prices and slower response occasions.

Parker cited a Southeast Asian chat assistant for example of Gemini Flash’s utility, whereas Gemini Professional excels in situations requiring complicated decision-making.

Google’s AI technique extends to Vertex AI, its enterprise developer platform that helps various fashions. Partnerships with Anthropic, Cohere, Mistral, and Meta’s Llama additional underline Google’s dedication to flexibility, providing enterprises a set of AI instruments tailor-made to their wants.

Enterprise adoption of AI is accelerating globally, regardless of considerations about return on funding (ROI). Parker noticed that whereas US enterprises concentrate on extracting worth from enterprise platforms, nations like India and China—historically developer-centric—are more and more embracing higher-level providers. “Constructing agentic methods in low-code environments is rising because the quickest technique to drive enterprise worth,” he stated.

Success tales similar to Apollo 24/7 Well being’s AI-powered session assistant spotlight the transformative potential of those options. These methods streamline processes and improve consumer experiences whereas driving down prices.

Learn this | How AI, GenAI malware is redefining cyber threats and strengthening the arms of criminals

“Over the subsequent 12–24 months, we’ll see progress in fashions and platforms, with enterprise worth shifting past builders to packaged options like agent-based platforms,” Parker stated, citing AgentSpace’s seamless integration with instruments like ServiceNow, Workday, and Salesforce.

Tackling rising competitors

Nevertheless, competitors on this house is heating up within the cloud infrastructure providers house, which incorporates infrastructure-as-a-service (IaaS), platform-as-a-service (PaaS) and hosted personal cloud providers. Third-quarter enterprise spending on cloud infrastructure providers was $84 billion worldwide, up $15.7 billion or 23% from the third quarter of 2023, based on the Synergy Analysis Group, with generative AI being a significant factor behind the market acceleration.

That stated, Amazon maintains a robust lead with a 31% market share, adopted by Microsoft (20%) and Google (13%). Among the many tier two cloud suppliers, these with the very best year-on-year development charges embrace Oracle, Huawei, Snowflake and Cloudflare, based on Synergy Analysis.

Parker countered, “Google’s vertically built-in stack—from silicon to infrastructure—gives a singular benefit, permitting us to optimize for value, latency, accuracy, and efficiency. This method permits tailor-made options throughout various use instances, similar to leveraging lengthy context home windows (as much as 2M tokens in Flash)”.

Google, based on Parker, anticipates a shift towards AI-first tech stacks, “which can redefine the hyperscaler (massive cloud providers suppliers) panorama as firms prioritize AI partnerships over conventional infrastructure”.

He added that Google has made “important enhancements over the previous yr, lowering errors and enhancing reliability”. For regulated environments, human oversight stays essential. Instruments like grounding with Google Search improve accuracy, significantly for forex and factual checks, “though no mannequin is flawless but”, Parker asserted.

As an worker myself, I exploit instruments like our Gemini system day by day, showcasing the significance of adopting superior applied sciences.

Google, he added, can be championing power efficiency–from water cooling information centres to optimizing tensor processing items (TPUs), “sustaining management in sustainability whereas assembly AI compute calls for.”

Ongoing studying

However what abilities will staff want on this period of excessive automation?

“As an worker myself, I exploit instruments like our Gemini system day by day, showcasing the significance of adopting superior applied sciences,” Parker stated. He acknowledged that skilling the workforce is important—not only for technical groups however throughout the board.

Parker additionally underscored the significance of multimodal AI, like the corporate’s Gemini fashions, which combine textual content, speech, photos, and video. “Over 80% of latest conferences highlighted the rising demand for such capabilities, which open up new use instances and solidify AI’s function in enhancing human productiveness,” he stated.

Coaching, based on Parker, is not nearly constructing AI but in addition utilizing it successfully. “As an example, in 2023, we dedicated to coaching 1 million folks in two years, however we achieved that aim inside 12 months by leveraging platforms like YouTube and focusing on various audiences—startups, conventional employees, and past,” he added.

Additionally learn | 5 synthetic intelligence tendencies to trace in 2025

Acknowledging that the tempo of change is difficult, Parker stated Google is adapting by quickly growing content material and utilizing its huge assets to distribute it. Schooling helps folks see AI as an financial profit somewhat than a menace to their abilities, he concluded.