Cresta Launches Deputy Agent for Post-Call Follow-Up Work

What happened

Cresta, the AI-for-contact-centres vendor, has launched a new product called Deputy Agent, designed to continue working on a customer case after the live interaction has ended. Rather than stopping once an agent hangs up, the AI agent is built to carry on handling the follow-up work that a human conversation typically leaves behind.

Cresta frames Deputy Agent as an extension of its existing AI agent assist and contact-centre automation portfolio, aimed squarely at the post-interaction gap that most customer service technology has historically ignored.

Why it matters

Most contact-centre AI to date has focused on the moment of conversation itself — real-time prompts, sentiment detection, call summarisation. Deputy Agent signals a shift in where AI value is expected to come from next: the unglamorous tail-end of a service interaction, where case notes, follow-up actions and system updates get completed, delayed or dropped entirely. By extending agentic AI into that post-call window, Cresta is positioning its platform as infrastructure for end-to-end case resolution, not just conversational support.

For digital transformation and operations leaders, this matters because it reframes contact-centre automation as a workflow problem, not just a conversation problem. Closing the gap between "call ended" and "case actually resolved" has direct implications for handle time, rework, and the operational debt that accumulates when follow-through is inconsistent.

The Renascence take

The interesting part of this launch isn't the conversation AI — it's the admission that service quality is mostly decided after the customer has stopped paying attention. That's where most experience debt quietly accumulates: promised callbacks that never happen, notes that never reach the next agent, actions that get queued and forgotten.

Customers judge an interaction as "resolved" long before the organisation has actually finished the work behind it — and that gap is where trust erodes silently, without complaints, until the issue resurfaces. An AI agent that keeps working after the hangup is really a bet that most service failure is a workflow failure, not a conversation failure. Operators evaluating this kind of tool should ask not "does it sound human?" but "does it close loops a human reliably forgets to close?" — and measure it on downstream recontact and rework rates, not on call-time metrics alone.