AI Agents Just Killed the Multimillion-Dollar IT Migration Contract

ZURICH and SUNNYVALE, Calif. – October 08, 2026

In the strategic landscape of 2026, the intersection of technical execution and global business strategy has never been more critical. For years, enterprise technology leaders have accepted a painful reality: modernizing core infrastructure is a slow, expensive, and highly disruptive process. But a quiet revolution in how technical work gets done is now challenging that status quo. Today, premium sportswear brand On, in partnership with Google Cloud, announced a milestone that fundamentally rewrites the economics of enterprise IT.

By deploying specialized AI agents, On successfully migrated 24 core microservices to Google Cloud, compressing the typical three-month transition timeline down to just two weeks per service. Even more striking, a small internal team of engineers executed 15 of these migrations entirely in-house. They achieved this with under five minutes of planned downtime per service across production environments. This is not merely a technical upgrade; it is a blueprint for the "Agentic Era" of enterprise operations, proving that multi-agent architectures can handle complex, end-to-end cloud migrations that traditionally required multimillion-dollar contracts and armies of external consultants.

Disrupting the $500,000 Statement of Work

To understand the magnitude of this shift, one must look at the historical context of enterprise cloud migrations. Industry benchmarks have long shown that the average enterprise cloud migration takes approximately eight months and costs anywhere between $1.2 million and $4.5 million. Large-scale transformations can drag on for years—famously, Capital One's migration of its applications to the cloud took the better part of a decade. Furthermore, nearly a third of all cloud migrations miss their planned timelines, bogged down by the sheer mechanical friction of moving deeply integrated systems.

When On began scaling its global operations, the Swiss-born brand realized its technology foundation needed to evolve to support its next phase of rapid growth. The goal was to reduce fragmentation between compute, data, and AI, simplifying the architecture into a unified platform. However, migrating complex, deeply integrated microservices is notoriously difficult. Microservices architectures, while agile, can introduce significant operational overhead, and untangling them for a migration often leads to increased latency and a high risk of failure.

Faced with this complexity, On initially sought external support. One third-party vendor submitted a $500,000 quote to migrate just a fraction of the necessary microservices using traditional, manual methods. Instead of signing that statement of work, On pivoted. By leveraging an agent-led approach, the internal engineering team was able to complete the broader project far faster, completely bypassing the need for outside implementation resources.

"AI changes the economics of execution. We're seeing that firsthand when our engineers used AI agents to deliver a major cloud migration with no additional resources and zero business disruption," said Narek Verdian, chief technology officer at On. "The real opportunity is rethinking how technology work gets done, while ensuring human engineers remain accountable for the decisions that matter."

This development poses a severe macroeconomic disruption risk to global IT systems integrators and consulting firms. If a small, internal engineering team can utilize AI agents to map complex deployment patterns across more than 20 code repositories in hours instead of days, the traditional consulting business model—built on billing thousands of hours for manual configuration and replication tasks—faces an existential threat. The constraints have shifted, and the premium is no longer on raw human headcount, but on the strategic orchestration of intelligent agents.

The Blueprinted Migration: Anatomy of a Multi-Agent DevOps Pipeline

The success of On's initiative lies in the sophisticated deployment of a coordinated multi-agent system. Moving beyond the standard developer assistance tools that simply autocomplete code, On and Google Cloud deployed domain-specific AI agents capable of handling complex migration tasks from end to end.

The technical division of labor was meticulously structured. The multi-agent system divided the migration into specialized, automated steps. Certain agents were tasked with codebase analysis, parsing through dozens of repositories to identify dependencies. Others generated the necessary infrastructure-as-code configurations, while parallel agents created translation pipelines and executed testing validation. By offloading these highly mechanical and repetitive tasks, On protected its product engineering roadmap, allowing human developers to focus on strategic feature building rather than tedious plumbing.

Crucially, this autonomous execution was strictly bounded by human oversight. In enterprise IT, the cost of downtime is astronomical. Recent industry surveys indicate that a vast majority of mid-size enterprises report losses exceeding $300,000 per hour of downtime, while large enterprises can face costs ranging from $1 million to $5 million per hour. Given these stakes, handing the keys entirely to an AI system is a non-starter for any responsible chief information officer.

On solved this by implementing defined tasks with clear checks. Engineers reviewed the AI-generated code, ran the infrastructure changes in secure staging environments, and personally authorized the final production cutovers. Consequential changes remained strictly subject to human approval. This "human-in-the-loop" governance model is what allowed On to keep its planned production cutover downtime to under five minutes per service—a staggering achievement that practically eliminates the operational risk typically associated with core infrastructure changes.

"True agentic transformation requires a unified data and compute foundation," said Karthik Narain, chief product and business officer at Google Cloud. "On proved that specialized AI agents can fundamentally shift the efficiency and velocity of complex cloud migrations, setting the stage to modernize their critical data platforms on Google Cloud."

From Microservices to Office Workflows: A Full-Stack AI Makeover

The migration of its core services is only the foundational layer of On's broader corporate transformation. By establishing Google Cloud as its enterprise AI backbone, the company is now positioned to leverage this unified data and compute environment across its entire global workforce.

Modern retail and consumer brands are increasingly recognizing that IT infrastructure is not merely a back-office utility, but a strategic growth engine. With its data platforms modernized, On has rolled out Gemini Enterprise to every employee within the organization. This deployment shifts the power of AI from the exclusive domain of software engineers to the everyday workflows of non-technical staff.

Teams across the company are now empowered to build custom, no-code agents tailored to their specific operational needs. From automating executive reporting and synthesizing market research to streamlining scheduling and accelerating new employee onboarding, the integration of Gemini Enterprise is driving tangible business impact. Furthermore, On has positioned itself as a strategic partner and trusted tester for Google Cloud's single universal agent for work, providing critical feedback that will shape the future of enterprise software.

This comprehensive approach highlights a vital lesson for the 2026 business landscape. The organizations that will thrive are those that rethink work from the ground up, rather than simply automating existing, inefficient processes. By utilizing a reusable agentic blueprint, On has not only solved an immediate engineering challenge but has fundamentally rewired its organizational DNA. The brand known for disrupting the footwear industry with its proprietary innovations is now proving that the most profound performance enhancements are happening in the cloud, driven by AI agents and guided by human ingenuity.