With GenAI and Google Cloud, the future starts now
Google Cloud Next ’24 delivers on the promise of generative AI
The 2024 Google Cloud Next conference wrapped up earlier this month, a little less than eight months since the ’23 conference introduced us to a raft of innovations around generative AI (GenAI) and Google Cloud’s mission to get GenAI into everyone’s hands. Back then, the focus was on experimen-tation and discovery, but this year we saw how the learnings and knowledge from that phase have been put to work. Google and Alphabet CEO, Sundar Pichai, summed it up best during his video introduction on day one of the conference: “Last summer, we were just beginning to imagine how this technology could transform businesses, and today, that transformation is well underway.”
His point was driven home by wave after wave of examples and testimonials, with dozens of customers (including Slalom customers) from industries such as retail, healthcare, finance, and education illustrating the promise of GenAI made reality. Though it wasn’t stated in so many words, the message was clear: if you’re not already capitalizing on GenAI, now is the time to get started.
Firming the foundation
How is Google Cloud uniquely positioned to help organizations achieve the kind of GenAI success seen in the testimonials? It starts with a strong foundation. According to Amin Vahdat, VP/GM ML, Systems and Cloud AI for Google Cloud, the use of large language models (LLMs) has resulted in a 10x spike in the need for computing power per year. Google Cloud’s AI Hypercomputer architecture is addressing this need with advancements in its one-of-a-kind combination of hardware and software that optimizes compute, networking, and storage infrastructure for the GenAI age.
Only with an architecture foundation like Google Cloud’s can organizations achieve the performance promised by Gemini 1.5 Pro, Google Cloud’s multi-modal (text, video, audio, code) AI model that was purpose-built for scaling across a wide range of tasks. This latest version of Gemini can support up to one million tokens and process a full hour of video, 11 hours of audio, more than 30K lines of code, and in excess of 700K words.
This allows Gemini to have a broader and more detailed context, which results in more sophisticated reasoning and better problem-solving capabilities. What’s more, Gemini’s responses can be securely grounded with Google search and organizations’ internal data to improve response quality and reduce hallucinations.
One example of how Gemini can be used was on display in Slalom’s booth at Next–much to the delight of anyone who stopped by. Using Gemini 1.5 Pro and Google Distributed Cloud technologies, Spot the robot dog helps us illustrate use cases such as quality assurance (QA) for manufacturing and inventory management. For example, reducing latency could be very valuable for inspections focused on defect analysis to augment QA and quality control functions. With the combination of intelligence and mobility, AI robotic technology can help manufacturers reduce costs by going to the material instead of making the material come to a sensor. It can also help employees, for example by keeping them safe (going to areas that are inaccessible or risky for humans), engaged, and fulfilled (offloading repetitive tasks, shifting time to more meaningful work).
Putting AI to work
With Google Cloud’s AI models and infrastructure forming that robust, performant, and powerful foundation, organizations can take advantage of a wide range of GenAI capabilities in the form of agents. Google Cloud’s agents are designed to go beyond basic chat. They aim to independently execute tasks across applications, acting as an extension of the user, who enters prompts using natural language. Whereas chatbots focus primarily on dialogue and are trained on scripted conversations so they can answer a more predictable set of questions, AI agents use GenAI, LLMs, and natural language processing as the basis for their ability to reason. It’s this ability that allows agents to gain a deeper understanding of user intent, context, and the environment they operate in, which in turn enables them to suggest actions, anticipate needs, or execute tasks without explicit prompts, unlike traditional chatbots that generally have a reactive model.
Thomas Kurian, CEO, Google Cloud, broke down Google Cloud’s agent capabilities as follows:
- Customer agents: Using the Vertex AI Agent Builder, organizations can build agents that deliver human-like conversations and interactions for their customers. Because these interactions are grounded in the enterprise’s data, they remain on message and aligned to organizations’ guidelines for serving their customers.
- Employee agents: Using the example of an annual benefits enrollment process, this demo illustrated how an agent can boost productivity and collaboration and help employees navigate the organization’s policies and processes.
- Creative agents: With Gemini powering the upcoming Google Vids offering, Google Workspace customers will have access to an AI assistant for video writing, production, and editing.
- Data agents: According to Brad Calder, GM & VP, Google Cloud, data agents allow users to unlock new ways to find and act on meaningful signals from their data. Gemini and BigQuery make it easy to bring enterprise data into Google Cloud to help with data preparation, discovery, analysis, and governance.
- Code agents: Google’s internal use of Gemini Code Assist resulted in 40% faster completion time for development tasks and 50–55% less time needed for understanding new code, writing new code, and writing unit tests.
- Security agents: Protect data privacy, security, models, and agents from cyberattacks by helping SecOps teams detect, respond to, and prevent threats more effectively. The power of Gemini is now being offered in …
- Threat Intelligence: Use natural language prompts to get deep insights about threat actors.
- Security Operations: Summarize and explain findings, recommend next steps, and write and execute remediation playbooks.
- Security Command Center: Evaluate your security posture and summarize attack paths to help remediate risks.
Are you GenAI-ready?
Google Cloud’s big message at Next ’24 was this: GenAI is here–now, today–and organizations need to take advantage of its incredible capabilities. However, many firms are struggling to place themselves into that early mover position to take advantage of generative AI.
Slalom can help you identify and mitigate any friction or obstacles to your GenAI adoption before they happen so you can implement GenAI securely and in a way that meets your unique needs. We provide data and AI solutions that improve end-customer experience, reduce operational cost, and fuel innovation. Our AI Value Calculator advisory tool and methodology empowers organizations to realize the full potential of their AI investments by forecasting the total cost of ownership (TCO) and return on investment (ROI) while also looking into use case value comparisons and investment prioritization.
Slalom’s approach to guiding organizations towards a clear strategy for quantifying AI ROI–a vital metric to have before embarking on a GenAI program–sets us apart from partners that rush their customers into the build phase without being aware of the costs and potential return. We can also help you uncover AI’s opportunity across your teams’ roles and tasks with our enhanceIQ accelerator, which empowers leaders to make data-driven decisions on tech investments, training, and workforce readiness for a more adaptable workforce.
When it comes to GenAI, Slalom offers guidance and expertise on everything from organizing and managing your data for improved outcomes to providing more comprehensive service to your customers. Learn the many ways Slalom can help you succeed with Google Cloud and GenAI.
