Zoom Virtual Agent 3.0 was unveiled on Feb 25, as Zoom Communications, Inc. introduced new agentic AI capabilities designed to automate end-to-end customer resolution across voice and chat.
The latest version of Zoom Virtual Agent 3.0 (ZVA) introduces a new execution architecture that enables multi-step workflows across enterprise systems, with built-in governance and transparency. The company said the update aims to reduce repeat contacts and improve first-contact resolution as organisations seek to scale automation.
According to a Morning Consult report commissioned by Zoom, 43 per cent of surveyed consumers said chatbots fail to resolve their issues, while 38 per cent cited getting stuck in loops and 37 per cent said they had to repeat information.
Automation Beyond Conversational Containment
Zoom said enterprises are entering what it calls a “resolution economy”, where performance is measured by end-to-end issue resolution rather than containment rates alone.
“Agentic AI was just the beginning,” said Chris Morrissey, general manager of Zoom CX. “Zoom Virtual Agent 3.0 orchestrates multi-step workflows across systems, continuously learns from human resolutions, and provides full transparency into every agentic action.”
Built on Zoom AI Companion 3.0 architecture, Zoom Virtual Agent 3.0 can run workflows across compatible customer relationship management (CRM), billing and order management systems. Administrators are able to audit data sources, decision logic and workflow paths behind automated actions, the company said.
The platform also integrates with Zoom Contact Center, enabling it to extract insights from escalated cases resolved by human agents and apply validated recommendations to similar future queries.
Multimodal AI And Proactive Outreach
Further enhancements expected to be generally available in Spring 2026 include multimodal large language model (LLM) capabilities. These will allow the virtual agent to interpret customer-submitted images, documents and structured identifiers such as serial numbers.
Zoom said this could automate workflows that previously required manual review, including warranty claims. In one example shared by the company, the system can authenticate users, extract serial numbers from uploaded images, validate eligibility across backend systems, schedule device pickup and initiate replacement orders within a single interaction.
If escalation is required, the full workflow history is transferred to a live agent, reducing the need for customers to repeat information.
The company also plans to introduce proactive outbound engagement, allowing the virtual agent to initiate contact and complete tasks based on known events.
Zoom said it has already implemented the latest updates internally. Its no-match rate — where the virtual agent fails to understand user input — dropped from 35 per cent to zero, according to the company.
On its billing team, deflection rates rose from zero to 30 per cent in three months, saving more than 1,000 agent hours per month.
The launch comes as enterprises across Asia-Pacific increase investment in AI-driven customer experience (CX) tools to manage rising service volumes and cost pressures. Analysts have noted that contact centres in Singapore and the wider region are prioritising measurable outcomes such as resolution rates and operational efficiency.
