How Brasil Antenado replaced most of its call center with AI agents on WhatsApp, RCS and SMS — saving R$150 million in one year.
The program serves millions of citizens a month, nationwide. That used to require more than 2,000 human service seats. Volume was growing faster than the call center budget.
Agentic service across messaging channels, with AI agents that read the audio and images citizens send and qualify each case on their own. The human team kept the exceptions.
Citizens who reach out on their own and get an immediate answer, at any hour.
Proactive contacts for communication, guidance and follow-up, spread across the three channels.
Conversations started each month across the three channels, under the same service rules.
Human service seats carried every citizen contact, with cost and capacity bound by call center scale.
Reduction in human service within three months, with the remaining team focused on exception cases.
Citizens send whatever is at hand: a voice note, a photo of the equipment, a short message. The agent has to understand all three.
They carry the conversation from start to outcome, in plain language, with the same rules across all three channels.
WhatsApp voice notes are transcribed and handled as text — people who prefer to speak are not left out.
Photos sent by the citizen are read by AI to qualify the case and shorten the path to a solution.
Goatable orchestrates the operation across the three messaging channels and fires the proactive contacts; CX Agents run the conversation with audio and image reading; Macro Insights tracks volume, outcome and cost per contact.
We map the operating design, the channels involved and the projected cost per contact for your scenario.