March 23, 2026
Cresta Launches Knowledge Agent to Eliminate Contact Centre Guesswork
Cresta has launched Knowledge Agent, an AI assistant that monitors live contact centre conversations in real time and surfaces precise answers to agents without being prompted.
The tool operates as a browser sidebar, running in the background as an agent moves between applications during a call. It processes both what is being said on the call and what is on the agent’s screen, pulling in contextual data such as a customer’s loyalty tier, account status, or order history. Based on these, the assistant generates targeted answers and links to relevant source material. Rather than requiring agents to type a query into a search bar mid-conversation, Knowledge Agent surfaces information at the moment it is needed.
Ping Wu, CEO of Cresta, said: “Contact centre agents have a hard job. They’re constantly juggling multiple tools to find the information they need, all while speaking to the customer at the same time. Knowledge Agent is a second brain for contact centre agents. It hears what they hear and sees what they see, handling all of the searching for them and giving them the answers they need, right when they need them.”
The Toggle-Tax
Enterprise contact centres typically require agents to move between CRM platforms, internal knowledge bases, billing systems, and help desk tools within a single interaction. Cresta refers to the productivity drain caused by this constant switching as the “toggle-tax”, the cumulative cost of mental fatigue and lost time every time an agent leaves one system to find information in another.
Knowledge Agent follows the agent across browser tabs, so, for example, when an agent navigates from a CRM to a billing tool, the sidebar travels with them, maintaining access to relevant knowledge throughout. The tool uses what Cresta calls Context Fields to read specific data points from the screen and incorporate them into each response, replacing generic FAQ articles with answers calibrated to the individual customer’s situation.
A significant share of contact centre organisations are not providing enough ongoing coaching and support to help agents work through AI-driven workflows, a problem that tools like Knowledge Agent are able to solve by giving agents immediate, accurate guidance rather than relying solely on training and memory.
Intended Outcomes
Cresta says the tool lifts three specific metrics: employee satisfaction, first call resolution rates, and new agent ramp time. On the last point, Wu added: “With Knowledge Agent, generalists can handle a wider range of issues without transferring the customer or putting them on hold. And without the need to multi-task to find the information they need, they can focus on what they do best: serving the customer.”
Faster onboarding is a meaningful benefit for contact centres that face high agent turnover. When new hires can access precise, context-aware answers from day one, the knowledge gap between new and experienced agents narrows considerably.
Knowledge Agent is the first in a series of agentic assistants Cresta has planned for the contact centre environment. The company says its platform serves both human and AI agents within the same operation.
The Human Still Wins
The global call centre AI market was valued at $2.41 billion in 2025 and is projected to reach $13.52 billion by 2034, growing at a compound annual growth rate of nearly 21%. The direction of that growth is increasingly towards human augmentation rather than full automation.
Analysts predict that AI and conversational tools will save $80 billion in contact centre labour costs by 2026, but 75% of customers still prefer human agents for complex, sensitive, or emotionally driven issues. This pushes vendors towards tools that make agents faster and better-informed rather than replacing them.
With vendors increasingly competing on the quality of real-time intelligence they can put in front of human agents at the moment of need, the question is no longer whether AI will be present in the contact centre, but how effectively it can deliver relevant information without adding to agent workload.
