CVC Corp

Objective

Recover demand that shows up when the physical network is closed. CVC Corp sells travel through a network of stores whose value rests on the consultant, but whose opening hours let every intent expressed in the evening or overnight walk away. Lia occupies that window: it answers, understands the project, qualifies the lead, and the consultant finds a prepared file in the morning rather than a cold form. The commercial bet is that in-store conversion rises because the customer has already moved forward, not because the salesperson works faster.

The deployment

CVC Corp is Brazil's main travel distribution network, built on a footprint of physical stores. Since early 2026 the company has run Lia, a virtual agent that takes over customer channels when the stores are closed, roughly from 10 p.m. to 10 a.m. according to management's description on the earnings call. Lia talks to the customer, understands the travel project, qualifies the lead, and passes that work to consultants at opening. The mechanism is therefore not end-to-end automated selling: it is an overnight relay whose output is a better prepared sales conversation. The figures given at the second quarter 2026 results disclosure on 13 August 2026 cover the previous month: 1.3 million interactions between Lia and customers, and 50% engagement, management specifying that half of the customers who interacted with Lia went on to engage with a consultant in store. The executive further states that conversion has grown 150% compared with the period before Lia. The quarter's financial context is difficult on other grounds: net revenue fell 6.5% to 319.5 million reais and the adjusted net loss widened to 51.3 million reais, driven notably by higher airfares. What the case does not say: the exact baseline behind the conversion figure, the scope of channels Lia covers, and how much of the network is involved.

Results Proof B

The figures are those given by CVC Corp management at the second quarter 2026 results disclosure on 13 August 2026, which is a level A event by nature. The level retained is B because the only source reached carrying the statements about Lia is a secondary transcript of the earnings call (Investing.com), the company's primary document not having been obtained at collection. The figures do measure the system itself (interactions with the agent, engagement of customers who went through it, conversion before and after it went live) and not company context: the quarter's financial results, which are down, are cited as context and are not attributed to AI by any source, either favorably or unfavorably. Honest caveat on the most striking figure: the 150% increase in conversion is an executive claim with no published baseline or measurement method, and it is reported as such rather than as an independently measured result.

The internal detail is not public. Here is a proven approach that leads to the same result, to adapt to your stack.

1. 1
   Contact outside opening hours customer

   The customer reaches out to CVC when the stores are closed and enters a conversation with Lia.

2. 2
   Qualifying the project AI

   Lia talks through the travel project and qualifies the lead.

3. 3
   Building the file AI

   The collected elements are structured and attached to the customer for in-store pickup.

4. 4
   Handover store consultant

   At opening, the consultant picks up a file already qualified rather than a cold contact.

5. 5
   In-store conversion store consultant

   The consultant closes the sale; post-Lia engagement is tracked as the system's metric.