Inside Agentforce Contact Center: The New “AI Plus Human” Workforce

Salesforce SVP Gautam Vasudev explains why cost pressure is pushing adoption, but why cost-only thinking can backfire fast.

AI & Automation in CXInterview

Published: March 26, 2026

Rob Wilkinson

If you are rolling out AI agents in the contact centre, the big risk is not whether you automate enough, but whether you automate the right things. In this CX Today interview, Salesforce SVP Gautam Vasudev explains why cost pressure is pushing adoption, but why cost-only thinking can backfire fast. If an AI agent fails, customers escalate to a human anyway, and you end up paying twice.

Vasudev outlines where AI works best today: high-volume, low-risk, repetitive requests like order status and simple knowledge questions. But he is equally clear on where customers still want a person: complex, multi-step issues and high-stakes scenarios, especially in healthcare and financial services, where reassurance matters as much as speed.

The takeaway for CX leaders is practical. Treat AI agents as part of the workforce, tune them like you would any team member, and measure success against the full set of experience outcomes, including cost, customer satisfaction (CSAT), and revenue impact. Just as importantly, keep escalation to a human discoverable so customers stay in control when it counts.

Agent AssistAgentic AIAgentic AI in Customer Service​AI AgentsAutonomous Agents
Featured

Share This Post