Imagine a world where intelligent agents from different applications talk to each other through rich, dynamic conversations.
A world where your weather app knows you’re traveling to a rainy city, and your favorite shopping app proactively recommends and orders the perfect lightweight jacket so it’s waiting for you at a store near your hotel.
Where your analytics tool notices a drop in user engagement and automatically pings your marketing platform to launch a targeted re-engagement campaign.
And yes – even a world where software errors are not just detected, but diagnosed and fixed – without you ever opening a ticket.
This is a glimpse into what’s possible. At Google Cloud Next, this vision took center stage with a conceptual demo powered by Pendo’s behavioral data, the Cursor agent’s code generation capabilities, and a network of collaborative AI agents—all built on the Agent-to-Agent (A2A) protocol. Pendo was proud to be featured as a launch partner at Google Next keynote by Google Cloud CEO Thomas Kurian, along with 50 leading industry partners, like SalesForce and Box.
This early exploration positions us—and our customers—for a future where intelligent systems don’t just coexist, but truly collaborate.
What is Google’s A2A protocol?
The Agent-to-Agent (A2A) protocol is a standardized communication framework that allows software agents—think intelligent assistants inside apps—to talk to each other, share context, and coordinate complex tasks. It’s a giant leap toward fully autonomous, interoperable AI ecosystems.
Key A2A Capabilities:
- Common language: A shared format for messages, actions, and data.
- Cross-system collaboration: Agents from different vendors can work together.
- Smart automation: Agents can initiate and complete workflows without human intervention.
Google’s A2A protocol works securely hand-in-hand with another rising standard called Model Context Protocol (MCP), developed by Anthropic, which is especially popular among developers using tools like Cursor. While MCP focuses on giving agents access to the right data and tools, A2A focuses on helping those agents communicate and collaborate. Together, they make it possible for smart software to work in harmony—sharing context, coordinating tasks, and delivering better outcomes with less human effort.
Why this matters for Pendo
Pendo’s inclusion in the Google Next announcement isn’t just about integration. It’s about showing what’s possible when you combine rich software usage data, real-time user signals, and autonomous agents.
Through A2A, we can unlock a new generation of experiences that are:
- Smarter: Agents can leverage behavioral insights to make better decisions.
- Faster: Workflows that once required human coordination now happen instantly.
- Seamless: Systems communicate behind the scenes, reducing friction for users.
At Pendo, we see Google’s A2A protocol as a foundational building block for the future of Software Experience Management (SXM). By enabling intelligent agents to collaborate across tools and systems, A2A helps bring the vision of self-healing applications to life. Imagine a world where user friction is automatically identified, the root cause is diagnosed, and fixes are implemented—without the user needing to file a ticket. This isn’t just automation—it’s a new level of software intelligence that actively improves the experience for every user, in real time. A2A enables this, turning SXM from a measurement tool into a dynamic system for optimization and action.
A glimpse into the future: automated bug resolution with A2A
Let’s bring this to life with a use case we presented at Google Next:
The problem: Bugs, frustration clicks, console errors—classic signs of user pain.
The solution: A fully automated, agent-powered workflow. Here’s how it works:
- Detecting trouble: Pendo captures signals like an anomaly in frustration clicks and errors.
- Escalating smartly: Once validated, an agent calls Google’s Gemini API to generate a detailed bug report.
- Enabling agent collaboration: The Pendo Agent sends a Slack message to notify the end-user: “We found and logged a bug.” It then hands off the task to the Cursor AI Agent, which understands the codebase and fixes the issue autonomously.
- Closing the loop: Once fixed, the Cursor Agent notifies Pendo, which sends a follow-up Slack message: “The issue has been resolved.”
All of this happens without a single ticket filed, meeting scheduled, or context switch required.
That’s the power of agents talking to agents.
Taking an AI milestone from possibility to reality
The Agent-to-Agent protocol is more than a technical milestone—it’s a step toward a smarter, more collaborative digital world. We’re thrilled to build that future alongside Google and other forward-thinking partners.
This is about giving Pendo customers the superpower to turn user signals into smart actions.
Pendo is proud to lead the way. Stay tuned—we’re just getting started.