Unleashing agentic AI: How PwC, Salesforce and Google Cloud are powering the next wave of intelligent transformation

AI agents are multiplying fast — but they’re not always working together. AI agents can now be developed as embedded features within platforms, as standalone applications, or as highly specialized agents built on proprietary or open-source software development kits (SDKs). However, businesses can sometimes find it challenging to enable these various agents in ways that help them communicate, collaborate and scale effectively across complex enterprise environments. The latest strategic moves from PwC, Salesforce and Google Cloud underscore this urgency, spotlighting the next frontier: AI that spans platforms and scales to your enterprise.

Businesses are not only reimagining the future of work thanks to augmented intelligence — we have now reached a pivotal moment where human ingenuity and AI can empower each other with great synergy. PwC is helping clients deploy intelligent agents that work together — grounded in proprietary data, designed for real-world outcomes, and orchestrated across cloud environments. With the right strategy, companies can design, build and operate a digital workforce guided by human oversight, which helps keep Responsible AI principles at the helm. As we look ahead to Google Cloud Next, here’s how AI agents driven by three powerhouses — Salesforce’s Agentforce, Google’s Agentspace, and PwC’s agent OS — are converging to help clients lead.

Salesforce and Google: Bringing Gemini into the Agentforce ecosystem

In February, Salesforce and Google announced a major expansion of their partnership: Google Gemini LLM will be integrated into Salesforce's AI stack, including Agentforce to deliver a more connected and intelligent data foundation for clients.

Integrating Gemini helps empower Agentforce with a native ability to ground Google search results and leverage longer context windows than other models, both of which can help provide better RAG (retrieval-augmented generation) results, improving speed and accuracy when answering queries. But while the tech is impressive, adoption isn't automatic. The real differentiator is execution — turning AI potential into business performance. That’s where PwC can come in.

We are advancing the combined power of Google Gemini and Agentforce by designing multi-platform solutions with our clients. For example, by leveraging the mathematical reasoning capabilities of the Gemini model, we can provide a trend analysis on critical performance data to forecast market movements and embed it directly in an agentic CRM flow to recommend the next leading actions for our clients’ purposes.

PwC’s agent OS: Architecting the agent ecosystem

One way PwC is leveraging Gemini, Agentforce and Agentspace is by connecting its agents to a broader ecosystem through its new operating system, PwC’s agent OS — the first business transformation-focused solution for orchestrating enterprise-grade AI agents across platforms like Salesforce, Google Cloud, Oracle, Workday and others. While you can create any agent in Google Agentspace or Salesforce Agentforce using Gemini models, PwC uses agent OS within of any cloud or large language model (LLM), to orchestrate an entire cross-platform workflow that is tailored to a client’s needs. Think of it as the switchboard behind intelligent agent deployments — enabling companies to connect home-grown agents, third-party SDKs, and platform-native agentic platform like Agentspace or Agentforce in a secure, scalable way that can be up to 10x faster than traditional methods.

Whether you’re using Salesforce Agentforce, Google Agentspace or a custom agent built on Google Vertex AI or another platform, PwC’s agent OS can help enhance governance and compliance through its integrated risk management and oversight frameworks, and support interoperability across your enterprise technology stack.

It’s the operating system layer that helps make AI agents work together.

PwC is leveraging agent OS within our own organization and with clients in different industries to orchestrate adaptive workflows across essential enterprise systems. For example, a retailer uses both Salesforce's Agentforce and Google Agentspace to deliver a better customer service experience for customers on their website. Agentforce quickly identifies customer intent from a web chat inquiry about an order's shipping status with the customer service agent. PwC's agent OS helps orchestrate the flow, seamlessly routing requests and managing communication between Salesforce and Google Cloud's specialized inventory and fulfillment agents. Google Cloud's agents promptly provide real-time data on product availability and shipping status from back-end systems, which PwC's agent OS then consolidates into a coherent response. Finally, Agentforce delivers the unified, personalized reply directly to the customer, resulting in a seamless, efficient customer experience powered by the combined strengths of Salesforce, PwC and Google.

Google Cloud AI Experience Zones: Strategy meets reality

To help bring the potential of agentic AI to life in the market, PwC launched Google Cloud AI Experience Zones — collaborative innovation hubs where clients can engage directly with AI-powered solutions, including industry-specific agents powered by Gemini, agents on Google Agentspace and other Google Cloud AI services.

Located in Bengaluru, Washington D.C. and San Francisco, these hands-on environments help allow clients to:

• Explore agentic AI use cases for healthcare, retail, telco, financial reporting, sustainability and more
• Test how AI agents perform in real-world business scenarios
• Co-innovate with PwC and Google Cloud experienced professionals
• Experience how Gemini-powered agents can integrate with broader ecosystems like Salesforce

These Experience Zones are more than demos — they’re testbeds for transformation, giving clients a safe space to experiment, validate ideas and move confidently toward scaled AI adoption.

Where to start: PwC’s AI activation path for Agentforce and Agentspace

If you're unsure how to begin or scale your Agentforce agent journey, PwC’s AI Activation Path for Agentforce and/or Agentspace can help you build momentum with confidence. In just five hours, clients co-develop high-value agent use cases aligned to strategic business goals. Within five weeks, they see those agents in action — with support from a dedicated PwC pod to help configure, customize and deliver real outcomes in Salesforce environments, Google environments, or both.

Clients can receive:

• Defined agent use cases for sales and service
• Experience mapping and data planning
• Pre-built industry use case libraries
• Light configuration and deployment in client sandbox environments
• Immediate insights into ROI and scalability

Why this matters at Google Cloud Next

With Gemini now embedded in Salesforce and accessible via tools that clients are already using, the opportunity to connect Google Cloud and Salesforce AI ecosystems is real. PwC’s agent OS acts as the connective layer—seamlessly integrating Salesforce Agentforce and Google Agentspace to orchestrate intelligent workflows across platforms. By leveraging PwC‘s agent OS, organizations can unlock a harmonized agent ecosystem where insights, actions, and outcomes flow freely — without being siloed by technology stacks. But successful execution requires:

• A cohesive strategy to align business and IT
• The right framework to orchestrate agents across platforms
• Clear governance, security and operational standards
• Rapid activation paths to demonstrate value and build trust
• Physical hubs to co-create and test what's possible

Join us at Google Cloud Next, where attendees can interact with various agents in action at PwC Booth #2874 or request an invite to our “Cloud & Cocktails” event at PwC's Experience Center in Mandalay Bay's Chapel. Visit our Google Cloud Next webpage to request access to our exclusive lounge and onsite sessions next week. Whether you’re exploring Salesforce Agentforce, Google Gemini, or others, we can help you activate AI with confidence, clarity and speed.

Let’s design your agentic AI roadmap.