Salesforce adds Testing Center to Agentforce for AI agents

Salesforce Unveils Agentforce Testing Center to Streamline AI Agent Lifecycle Management

Salesforce has introduced the Agentforce Testing Center, a suite of tools designed to help enterprises test, deploy, and monitor autonomous AI agents in a secure and controlled environment. These innovations aim to support businesses adopting agentic AI, a transformative approach that enables intelligent systems to reason, act, and execute tasks on behalf of employees and customers.

Agentforce Testing Center: A New Paradigm for AI Agent Deployment

The Agentforce Testing Center offers several key capabilities to help businesses confidently deploy AI agents without risking disruptions to live production systems:

AI-Generated Synthetic Testing: Using natural language instructions, the tool generates hundreds of synthetic customer interactions to simulate real-world scenarios. This enables teams to test AI agents’ decision-making processes, including topic and action selection, and refine instructions to improve performance.
Safe Testing with Sandboxes: Sandboxes for Agentforce and Data Cloud allow teams to test agents in isolated environments that replicate production data and configurations. This enables rigorous user acceptance testing (UAT) without jeopardizing business operations.
Monitoring and Observability: Integrated tools like the Einstein Trust Layer provide audit trails, feedback loops, and advanced analytics for continuous improvement. New features like Agentforce Analytics and Utterance Analysis allow granular insights into agent performance, adoption, and accuracy.

Supporting a Limitless Workforce

Adam Evans, EVP and GM for Salesforce AI Platform, emphasized the importance of these tools in accelerating the adoption of AI agents:

“Agentforce is helping businesses create a limitless workforce. To deliver this value fast, CIOs need new tools for testing and monitoring agentic systems. Salesforce is meeting the moment with Agentforce Testing Center, enabling companies to roll out trusted AI agents with no-code tools for testing, deploying, and monitoring in a secure, repeatable way.”

From Testing to Deployment

Once testing is complete, enterprises can seamlessly deploy their AI agents to production using Salesforce’s proprietary tools such as Change Sets, DevOps Center, and the Salesforce CLI. Additionally, the Digital Wallet feature offers transparent usage monitoring, allowing teams to track consumption and optimize resources throughout the AI development lifecycle.

Customer and Analyst Perspectives

Shree Reddy, CIO of PenFed, praised the potential of Agentforce and Data Cloud Sandboxes:

“By enabling rigorous pre-deployment testing, we can deliver faster, more accurate support and recommendations to our members, aligning with our commitment to financial well-being.”

Keith Kirkpatrick, Research Director at The Futurum Group, highlighted the broader implications:

“Salesforce is instilling confidence in AI adoption by testing hundreds of variations of agent interactions in parallel. These enhancements make it easier for businesses to pressure-test autonomous systems and ensure reliability.”

Availability

Agentforce Testing Center: Closed pilot now; general availability in December.
Sandboxes for Agentforce and Data Cloud: Available today.
Agentforce Analytics & Utterance Analysis: Available today.
DevOps Center, CLI, and Change Sets for Agentforce and Data Cloud: Available today.

With these tools, Salesforce solidifies its leadership in the agentic AI space, empowering enterprises to adopt AI systems with confidence and transform their operations at scale.

🔔🔔 Follow us on LinkedIn 🔔🔔