Cresta's Deputy Agent AI Handles Post-Call Contact Center Work

Cresta has launched Deputy Agent, a new AI agent built to continue handling work after a customer interaction ends. Rather than stopping when a call or chat concludes, the system is designed to carry on completing tasks tied to that conversation once the customer has disconnected.

The launch extends Cresta's contact-center AI portfolio, which has focused on real-time agent assistance during live customer interactions. Deputy Agent shifts the company's focus to the post-interaction phase of service work — the administrative and follow-up tasks that typically fall to human agents after they hang up.

Post-call work has long been one of the least visible, least efficient parts of contact-center operations. Agents routinely spend time after every interaction summarizing notes, updating records, triggering follow-up actions or closing out tickets — work that adds to handle time and contributes to burnout, but rarely gets the same optimization attention as the live conversation itself. An AI agent purpose-built to absorb that workload signals a shift in where AI investment in service operations is heading: from assisting the conversation to owning the administrative tail of it.

For leaders modernizing service operations, the significance isn't just efficiency. If AI can reliably take over post-interaction tasks, human agents get time back to handle more complex, judgment-heavy work — and supervisors gain a new layer of automation to manage, rather than just monitor. It also raises the bar on what "agent assist" means: the measure of a good AI deployment is no longer limited to what happens while the customer is on the line.

Most coverage of contact-center AI fixates on the moment of conversation — the chatbot, the real-time prompt, the sentiment score. The more interesting shift is happening after the customer leaves, in the unglamorous admin work that quietly shapes employee experience and service consistency.

Post-call work is where agent fatigue and data quality problems actually live, yet it gets almost none of the design attention lavished on the conversation itself. Automating it isn't just an efficiency play — it's a statement about what organizations think agents' time is worth. The operators who get this right will treat Deputy Agent-style automation as a redesign of the agent's job, not a bolt-on tool, and will watch closely for whether it actually reduces burnout or just shifts the bottleneck to someone reviewing the AI's work.