Google Cloud is forming a joint team with Accenture, sending engineers to enterprises to assist in deploying the AI tools and services. The competition among large models is shifting from model performance to the ability to be implemented in real-world applications. Whoever can integrate these tools into their business more quickly will have a better chance of turning their initial investments into revenue.

Train up to 1,000 engineers at most.

According to the arrangements between both parties, Google will train up to 1,000 Accenture engineers to support them in developing customized AI applications for enterprises based on Gemini Enterprise. These positions require not only an understanding of corporate processes but also the ability to integrate AI tools into existing systems, with the goal of shortening the deployment cycle.

Google Cloud still lags behind its main competitors

Although Google Cloud's revenue in the second quarter reached $24.8 billion, with a significant portion of that growth coming from its AI business, its share of corporate AI spending is still noticeably lagging behind. Data from August Ramp shows that Google accounts for about 6% of corporate AI spending in the United States, Anthropic accounts for 43.5%, and OpenAI accounts for 39.7%.

This means that Google not only needs to continue to invest GPU, data centers, and power resources, but also needs to prove that these investments can lead to more stable corporate demand. The report mentions that as of June 30, Google's procurement commitments and contractual obligations under Alphabet had reached a total of 811 billion US dollars, indicating that the scale of investment in AI infrastructure is still expanding.

Enterprise deployment remains a key bottleneck.

The current reality faced by the AI industry is that although companies continue to increase their AI budgets, the returns are not always clear. Many firms are not short of willingness to try new approaches, but the real challenge lies in how to integrate these models into their work processes and achieve sustained cost savings or revenue growth.

Google has expanded this deployment model several times this year. Previously, Google Cloud announced an investment of $750 million to build a partner ecosystem, embedding its own engineers into consulting firms such as Capgemini, Cognizant, and Deloitte. It also reached a multi-year partnership with CVC Capital Partners, deploying relevant engineers directly to their portfolio companies.

Consulting firms are also dealing with new competitors.

This type of on-site deployment model is not exclusive to large cloud providers. Reports mention that teams related to Anthropic, such as Ode, as well as The Deployment Co from OpenAI, are also expanding. On one hand, these newcomers help model companies compete for corporate clients, and on the other hand, they are also squeezing the service space of traditional consulting firms.

For Accenture, collaborating with Google is also part of a series of AI deployment projects this year. The company previously advanced similar projects with Microsoft in March, launched related plans with ServiceNow in May, and engaged in joint cooperation with SAP in June. As businesses pay more attention to whether AI investments will yield tangible results, deployment capabilities are becoming a new battleground for cloud vendors and consulting firms to compete over.