Google and Accenture Launch Joint Team to Put AI Engineers Inside Big Companies
Google holds just 6% of enterprise AI spending against rivals at nearly 40% each. A new unit training 1,000 dedicated engineers is its latest push to close the gap.
Key points
- Google Cloud and Accenture have formed the Accenture Gemini Enterprise Business Group, a joint unit that will embed engineers directly inside large companies to set up Google's AI tools.
- Google plans to train up to 1,000 Accenture engineers specifically to build custom AI applications on its Gemini platform.
- As of August 2026, Google accounts for roughly 6% of U.S. Enterprise AI spending, compared to Anthropic at 43.5% and OpenAI at 39.7%, according to financial software firm Ramp.
- The move follows a $750 million partner-ecosystem commitment Google Cloud announced earlier this year, placing Google's own engineers at Capgemini, Cognizant, Deloitte and other firms.
Google Cloud and Accenture, one of the world's largest technology consulting firms, have created a dedicated unit whose sole job is embedding engineers inside big companies to get Google's AI running. First reported by The Wall Street Journal, the unit is called the Accenture Gemini Enterprise Business Group.
Google will train up to 1,000 Accenture engineers to work directly with corporate clients, building custom applications on Gemini, Google's suite of large language models (the technology behind AI chatbots and assistants).
Why does any of this matter to ordinary people?
For anyone whose employer uses outside technology consultants, this is where the rubber meets the road. The hard part of enterprise AI has turned out to be making purchased systems actually work inside real businesses. These so-called forward-deployed engineers, or FDEs, are simply engineers who work on-site at a client rather than at the technology firm's offices, acting as long-term guides. They're meant to close the gap between "we bought some AI" and "the AI is saving us money."
We've tracked this FDE trend since our first story on the model on 15 July, and the competition for that on-site role has intensified sharply.
How far behind is Google?
Quite far. The Ramp data tells a blunt story.
| Company | Share of U.S. Enterprise AI spending (Aug 2026) |
|---|---|
| Anthropic | 43.5% |
| OpenAI | 39.7% |
| ~6% |
Alphabet, Google Cloud's parent, had accumulated $811 billion in purchase commitments and contractual obligations as of 30 June, a vast financial bet on AI demand growing fast enough to justify spending on chips and data centres. That return isn't materialising yet, which is exactly why generating enterprise demand matters so urgently.
The Accenture unit is Google's most direct attempt to create that demand. Earlier this year Google Cloud pledged $750 million toward embedding its own engineers across multiple consulting firms, and struck a separate deal to place engineers inside the portfolio companies of investment firm CVC Capital Partners.
What happens next?
The FDE model is a competitive arena of its own now. Accenture launched a similar engineer-placement practice with Microsoft in March, one with workflow software company ServiceNow in May, and a joint programme with business software firm SAP in June. It isn't waiting on Google alone.
Smaller specialist firms are also pressing in. Companies whose entire business is embedding engineers inside organisations to build bespoke AI workflows, such as Ode (which works with Anthropic) and OpenAI's own The Deployment Co., are putting pressure on the big consultancies at the same time.
For employees at large organisations, the practical upshot is that AI tools are likely to arrive with more hands-on setup help than before. Whether that translates into genuine productivity gains, or just a faster rollout of tools that still don't quite fit, will depend on how well those engineers understand the actual work.
