Salesforce has unveiled new observability tools named Agentforce 360 to help teams understand the performance of AI agents. As AI adoption grows, many organizations shift their focus from building agents to managing them effectively in production. According to Madhav Thattai, Salesforce’s SVP and COO of Agentforce, there has been a three-fold increase in implementations of AI agents in production, highlighting the need for robust observability tools.
The new tools encompass three main areas: Agent Analytics, Agent Optimization, and Agent Health Monitoring. Agent Analytics tracks agent performance over time, providing visibility into usage metrics and identifying ineffective areas. Agent Optimization focuses on diagnosing issues by tracing how agents make decisions and grouping requests to uncover patterns. Lastly, Agent Health Monitoring emphasizes uptime and responsiveness, providing alerts on critical errors and performance metrics.
For instance, 1-800Accountant has utilized these tools to streamline their operations during peak periods, highlighting how the observability tools enhance monitoring and allow for rapid iteration and improvement of AI agents. The tools are currently available, with plans for regional rollout in APAC. The overarching goal is to enable businesses to enhance client experiences by efficiently managing and improving their AI agents.
👉 Pročitaj original: CIO Magazine
