August 05, 2025
ASAPP Levels Up Contact Center AI with New Accuracy and Trust Tools
ASAPP has unveiled major updates to its flagship product, GenerativeAgent—an AI voice and chat agent designed to handle complex, multi-step conversations. With this expansion, ASAPP offers tools that strike the right balance between automation at scale and the human oversight needed for real-world customer service.
Built from the ground up to power fully autonomous interactions, GenerativeAgent integrates seamlessly with enterprise systems, including customer relationship management (CRM) platforms, CCaaS solutions, and internal APIs.
This deep integration enables the AI to deliver real-time resolutions informed by a business’s historical customer data, unlocking value instantly while maintaining high levels of precision and security.
“These new capabilities from GenerativeAgent equip CX leaders with the tools to safely and confidently scale automation and the ability to monitor, control, and continuously improve how AI agents perform over time,” said Devidas Desai, senior vice president of product management at ASAPP.
Among the new features is Human-in-the-Loop Agent (HILA) with Approver Mode, which introduces an optional layer of human review. Contact centre agents or supervisors can now monitor and approve AI-generated responses either in real time or after the fact, improving the system’s output over time while ensuring regulatory compliance and customer satisfaction.
In addition, ASAPP improves transparency and quality control through advanced monitoring and tuning tools, allowing teams to review conversations, flag anomalies, enforce guardrails, and gain critical insights from performance trends. This builds trust in AI automation while ensuring consistency with brand standards and service expectations.
To support safe roll-outs, the updated GenerativeAgent also includes a Testing and Simulation environment. This feature lets companies rigorously test AI behaviour in a sandbox before deploying updates into live customer interactions, reducing risk and ensuring the AI behaves as intended.
