Salesforce adds observability to Agentforce, aiming to boost AI performance and trust

Salesforce is expanding the capabilities of its Agentforce 360 Platform with new observability tools designed to give organizations real-time visibility into AI agent performance. The update, announced today, aims to help companies scale their use of AI agents with greater confidence, transparency, and control.

As AI-powered agents assume more customer-facing and operational roles, the need for continuous monitoring has become increasingly critical. According to Salesforce’s 2025 CIO Trends Report, AI adoption has increased by 282% year over year. However, scaling AI isn’t just about launching more bots — it’s about ensuring they behave as expected, improve over time, and deliver measurable value.

That’s where the new features come in. Delivered through Agentforce Studio and built on Salesforce’s Session Tracing Data Model, the observability suite gives teams deep insight into every AI agent’s actions — from how they respond in conversations to how reliably they perform behind the scenes.

The tools fall into three main categories:

• Agent Analytics: Provides a detailed view of agent usage and effectiveness, surfacing trends, KPIs, and underperforming flows. Teams can use these insights to make targeted performance improvements.
• Agent Optimization: Offers granular session tracing, helping teams understand how agents make decisions and identify areas where they may go off track. The goal is to enable faster debugging and smarter tuning.
• Health Monitoring: Tracks uptime, latency, and errors in real time, flagging silent failures and helping ensure agents stay responsive and available.

Taken together, these tools turn agent performance into something you can actually see, measure, and manage — a significant step forward for organizations trying to build what Salesforce calls “Agentic Enterprises.”

The goal is to promote more effective collaboration between human teams and AI agents by addressing blind spots and uncertainty. If businesses are going to trust AI with key parts of the customer journey, they need assurance that those systems are operating safely, ethically, and effectively.

For marketers, this kind of observability could be a game-changer. With AI agents increasingly handling tasks such as lead qualification, customer support, and campaign personalization, the ability to track their performance — and course-correct in real-time — will be essential for maintaining brand trust and delivering consistent customer experiences.

In short, Salesforce is betting that observability will be the next must-have capability in enterprise AI. And with companies accelerating their AI deployments, that bet may be well-timed.

The post Salesforce adds observability to Agentforce, aiming to boost AI performance and trust appeared first on MarTech.