Google intros generative AI consulting services and updates

Google's consulting services

Enterprises will also have access to generative AI consulting offerings from Google to help customers discover trends within their company data, summarize large amounts of information, automate business processes and create personalized content.

Google also introduced sample architectures and business-oriented workflows for enterprises looking to accelerate their entrance into the generative AI market.

The tech giant did not say when these services will be available.

Google moving swiftly into generative AI consulting services offerings is not surprising, said Daniel Newman, an analyst with Futurum Research.

"A lot of times, these tech companies have been doing some level of consulting, usually in partnership with systems and consulting firms," he said. "A lot of these [system integrators] work with everybody's technology. Google wants to make sure they control their destiny."

However, offering consulting services does not mean Google won't continue to partner with Deloitte, Accenture and other systems integrators, Newman said.

Meanwhile, the consulting offerings will support enterprise customers that are part of the generative AI wave. They are in the early days of understanding the rapidly developing large-language-model-based technology that lets organizations easily automate tasks, among other capabilities.

"Every enterprise right now is doing their due diligence to understand how they can safely, reliably and advantageously put these models to work," said Bradley Shimmin, an analyst at Omdia. "What I think Google is doing right now is trying to jumpstart that process for them."

Google is also trying to work on one of the key challenges it faces in the generative AI market, Shimmin said.

Compared to rival Microsoft, whose close alliance with GPT creator OpenAI has rocketed the tech giant to the top of the generative AI market, Google needs to expand its ecosystem of developers building capabilities with AI, he said.

Therefore, this is Google's attempt to build out that ecosystem and provide the tools -- whether training, professional support or tooling -- to start putting Google's generative AI PaLM API to work, Shimmin continued.

"Google has a number of pre-built solutions, but that's not going to account for more than a sliver of the broader market over time," he said. "That long tail for generative AI is going to be highly specialized within given verticals and use cases and departments."

Updating Vertex and targeting enterprise search

Besides its consulting services, Google also revealed that generative AI support in the Vertex AI-managed machine learning platform is now generally available.

With this, Model Garden -- an environment that lets users search, discover, and interact with Google's foundation models (soon also available with open source and third-party models) -- now includes access to more than 60 foundation models.

Every enterprise right now is doing their due diligence to understand how they can safely, reliably and advantageously put these models to work.

With the Vertex update, developers can access Google's text model powered by PaLM 2 as well as get text embeddings and tools in the vendor's Generative AI Studio for model tuning and deployment.

"It seems to me, with Vertex, they're making sure that the data sets are secure and private and separated," Newman said, referring to Google's claim that data, governance and safety features back Vertex AI.

Google, which has dominated consumer search for years, is also targeting enterprise search.

On June 6, the tech giant expanded access to Enterprise Search on Generative AI App Builder. The capability lets enterprises create custom chatbots and search engines that provide outputs from specific data sources and can handle multimodal data, such as images.

Users of the generative capabilities will have control over their data, even though the app builder is a fully managed service, Google said.

The vendor said enterprises such as Priceline and Vodafone are experimenting with Gen App Builder. Priceline is using the capability for internal search engines and for a new chatbot for consumers. Vodafone is using Gen App Builder to build a tool that can search documents and understand special commercial terms and conditions.

Enterprise Search on Gen App Builder is set up differently than other generative AI tools that OpenAI or Microsoft have opened up to developers, Shimmin said.

"The sort of understanding that Google is putting out there is that this isn't just a sandbox for developers to play in," he said. "This is a component of a broader toolchain that has, as its endpoints, something actually working in the company doing something, not just, 'Oh, that was an interesting finding.'"

Gen App Builder will be helpful for either an individual using it to automate some aspect of their daily job or a company inserting it as a capability into their existing chatbots, Shimmin continued.

Overall, Newman said that Google is looking to make it easier for enterprises to build and create their own generative AI capabilities.

"[It's] kind of inspiring in terms of what the potential is," Newman said, adding that it also feels scary to him that enterprises can begin ingesting their data and building apps with little effort. "It does feel like there's a very low barrier of entry. Having said that, I haven't seen anything that was actually built."

Proof that the new generative AI tools actually work will come when and if enterprises build some useful applications using them, he said.

Google also introduced Duet AI for Google Workplace Enterprise, a capability that enables collaboration within organizations. It is now available for preorder. Features include writing and refining content in Gmail and Google Docs, creating original images from text in Google Slides, and turning ideas and data into insights on Google Sheets.

Esther Ajao is a news writer covering artificial intelligence software and systems.