The Problem
The company was receiving a growing number of people interested in the destination, but most arrived full of questions and without enough clarity to move forward. The support team had to guide each person from scratch: explain products, answer questions about documentation, payment, accommodation, tours, and exchange programs — every single time.
The result was a real bottleneck: overloaded agents, leads dropping off mid-journey, and little time left to focus on people already ready to buy.
We identified three major pain points:
- Repetitive, non-scalable support: the same questions appeared hundreds of times, consuming team time with information that could be automated.
- No qualification at the top of the funnel: leads reached the human team with zero context — agents had to discover from scratch what each person wanted, their profile, and which product they were interested in.
- Limited reach: human support volume had a ceiling. Growing audience meant growing headcount, making operations expensive and hard to sustain.
Understanding the Agent
The agent was designed to act as the first point of contact: a friendly, positive, and trustworthy assistant with simple, warm, and approachable language. Its role went beyond answering questions: it guided each visitor through a consultative journey to understand the destination, identify the best-fit product, and arrive at the human team ready to decide.
To do this, the agent had access to a structured data base with information about all products, tours, payment options, required documents, and frequently asked questions. Instead of memorizing everything in fixed instructions, it queried these tables dynamically — making responses more accurate and allowing updates without rebuilding the agent from scratch.
A key experience design detail: when the agent detected that this was the user’s first international trip, it celebrated the moment. A small tone shift — more encouraging and welcoming — created emotional connection and confidence right when users needed it most. Small touches like this change how users perceive the entire conversation, and Caramelo AI is always focused on making support more human.
Recap:
- Identify user profile and intent at the beginning of the conversation, understanding whether the person is looking for exchange, tourism, expeditions, standalone tours, or accommodation.
- Present only relevant products for that profile, filtering the catalog and clearly explaining options without overwhelming the visitor.
- Answer questions accurately by consulting the knowledge base for documentation, payment, policies, and FAQs.
- Recognize emotional milestones, such as a first international trip, and adapt communication to reinforce trust and excitement.
- Capture the lead at the end of the conversation, collecting relevant information for the sales team.
From curious visitor to qualified lead
With the agent in operation, the company transformed top-of-funnel intake into a productive stage — not just an informational one. Visitors who previously arrived lost now followed a structured journey, with the right questions in the right order and the right products introduced at the right time.
The human team stopped being the first filter and started working where it matters most: conversion. Each lead handed over by the agent already included mapped profile, identified product interest, and resolved basic questions — significantly reducing average first-contact handling time.
On top of that, operations gained real scale. The agent can support any volume of simultaneous visitors without quality degradation — allowing the company to reach a much larger audience without proportional team growth. The result was a more efficient operation, a more consistent visitor experience, and a sales funnel with much better-prepared leads ready to take off.