Salesforce on Measuring Agentic Work Outcomes

What this episode covers

“Enterprises cannot vibe operate,” declares Madhav Thattai, GM and EVP of Salesforce AI, on this episode of Bloomberg Intelligence’s Tech Disruptors podcast. Thattai and BI senior technology analyst Anurag Rana discuss why Salesforce thinks the “SaaSpocalypse” narrative misses how agents change outcomes and how the company is trying to measure that shift with Agentic Work Units (AWUs) — a way to track work performed rather than just token inputs. Thattai outlines Salesforce’s four-part agentic enterprise framework: system of context, system of work, system of agency and system of engagement. The conversation explores how large-language model flexibility plus deterministic workflows are used to reduce “prompt doom loops,” improve latency and manage cost. The discussion also covers how Salesforce is monetizing agents via license bundles and consumption, and why AWUs and session-trace tooling are positioned as budgeting and operational aids for CIOs and CFOs.

PodParley-generated summary based on available episode metadata and transcript content.