The customer-service chatbot most people know is the one that asks you to rephrase the question three times and then hands you to a human. Zoom’s new Virtual Agent 3.0 (ZVA 3.0) is a pitch for the opposite: an AI agent that takes the problem in, works out what to do, and does it — without a person on the other end.
What changed in version 3
Three things stand out from Zoom’s announcement. First, the agent works over both voice and chat, so a phone call gets the same automation as a web widget. Second, it can read what customers send it — documents, forms, photographs — and make decisions based on their contents, which removes the step where a human has to open the attachment. Third, it runs end-to-end workflows: rather than answering a question and stopping, it can carry a task through the systems behind the desk, including CRM, billing and order management. Zoom says the release will reach customers in stages.
The metric that matters: first-contact resolution
Zoom is explicit that the target is “first contact resolution” — the share of issues fixed the first time a customer gets in touch, with no follow-up needed. It is the number support managers care about most, because every repeat contact costs money and goodwill. An agent that can verify a document and process a refund in one session moves that number in a way a scripted bot never could.
Zoom is not a video company any more
It is easy to miss how far Zoom has drifted from the pandemic-era meeting app. The company now sells itself as a cloud communications provider for businesses, and contact-centre software is one of its growth bets. Virtual Agent 3.0 competes less with Google Meet than with Salesforce, Zendesk and the wave of AI agent start-ups promising to automate the help desk.
What it means for Bangladesh
Bangladesh has spent a decade building a call-centre and BPO sector that sells human attention by the hour. Agents like this do not eliminate that work overnight — voice recognition in Bangla-accented English and regional languages remains a weak spot for every vendor — but they change the shape of it. The jobs that survive are the ones handling exceptions the bot cannot: the angry customer, the ambiguous claim, the edge case. Firms in the sector that train staff for judgement rather than scripts, and that learn to supervise AI agents rather than replace them, are the ones positioning for the next contract cycle.
The catch
An agent that acts autonomously also fails autonomously. A misread form becomes a wrong refund; a misunderstood voice call becomes a cancelled order. The companies deploying these tools will need audit trails and easy human override — the same lesson every industry is learning about agentic AI at once.
