SUCCESS STORY
In partnership with Cloud Humans, Insider implemented AI agents capable of automating support, anticipating logistical issues, and helping customers choose products in e-commerce, scaling operations without increasing the team.
Metric | Result |
|---|---|
Support cases resolved by AI | 76% |
Satisfaction in interactions with AI | 81% |
Increase in conversion with AI | +18% |
About Insider
Insider is a Brazilian brand positioned as a company that applies technology to clothing. Its products combine design and textile technology. The customer relationship is a fundamental part of the brand experience, making the CX operation a critical piece in sustaining the company's growth. Its customers frequently turn to support to clarify questions about products, sizes, exchanges, and orders, expecting quick, clear, and human responses.
Rapid growth, operational pressure and risks in CX
With Insider’s accelerated e-commerce growth, customer service began facing a challenge typical of expanding digital companies: keeping pace with the business without compromising the brand experience.
Title | Description |
|---|---|
Growing volume of interactions | As the e-commerce business grew, the number of questions about orders, exchanges, and products kept increasing, putting pressure on the support operation. |
Constant operational pressure | For the bot to work properly, the customer had to type exact terms. So it was necessary to register multiple variations of the same question, which made maintenance complex and inefficient for the volume of tickets the company received. |
Rising support costs | Scaling support only by adding staff would mean increasing operational costs and compromising the efficiency of the operation. |
Risk to the customer experience | In a digital model, where much of the relationship with the customer happens through support, any delay or inconsistency in service directly affects brand perception. |
In an operation where agility, clarity, and empathy are part of the customer’s perception of value, CX could not become just a support area. It needed to grow alongside the company, without relying on team expansion at the same pace. That meant dealing with more demand, greater operational pressure, and less room to build a truly consistent experience.
“Bad CX breaks retention. And bad CX breaks the brand. Anyone who says CX is a cost is looking at it the wrong way. CX is revenue when it is strategic.”
Cacau Pellegrino
Head of CX at Insider
The challenge, then, was not just to serve more customers. It was to create an operation capable of scaling efficiently, preserving the human touch at the right moments, and sustaining the company’s growth without turning customer service into a bottleneck for the business itself.
Insider + Cloud Humans: a strategic partnership
It was in this context that Insider began its partnership with Cloud Humans to redesign its customer service operation with the support of AI.
More than implementing a chatbot, the goal was to build an operation in which AI agents could absorb a large share of operational questions, maintaining the quality of the experience and freeing the human team for more complex interactions.
Main changes:
The operation began to rely on AI agents trained based on the support history and the real context of Insider's operation
AI began to act as the first support layer (N1), answering recurring questions and collecting information before escalating more complex cases.
The AI agents were integrated into the company's CX workflows, enabling responses based on operational data and customer context.
With continuous monitoring by the CX team, the AI began to evolve constantly based on real interactions with customers.
“Scaling is like changing a tire while the car is moving. You need to evolve, grow, innovate, and still keep up with the operation that already exists.”
Cacau Pellegrino
Head of CX at Insider
The “ClaudIA phases”
The implementation of AI at Insider was structured as an evolution journey. Instead of launching a complete automation all at once, they built the solution in stages of maturity.
Each stage expanded AI’s role within the operation, starting with learning from real interactions and evolving to the creation of an agent capable of also supporting the sales process.
Stage | Title | Description |
|---|---|---|
01 | Human learning | Before putting AI into production, they started by analyzing real conversations with customers. The team cataloged the main reasons for contact, identified time bottlenecks, and built a knowledge base with the brand’s tone of voice and vocabulary. This stage ensured that AI was trained based on the real context of the operation, not just theoretical support flows. |
02 | ClaudIA Respondona | The first version of ClaudIA went live to answer simple, recurring questions, such as questions about exchanges, delivery times, and order status. The team nicknamed this phase “ClaudIA Respondona” because the AI solved the cases correctly, but still needed to evolve in tone of voice and empathy. |
03 | Proactive ClaudIA | ClaudIA began identifying exceptions in the customer journey, such as logistics delays, and sending messages proactively, explaining what had happened and offering alternatives or compensation. This change reduced the volume of reactive interactions and increased customer trust in the service. |
04 | IAra, the sales agent | In the most advanced stage of the project, IAra was born, an AI agent designed to also operate in the sales process. She can answer questions about products, suggest item combinations, and support indecisive customers during the shopping journey. Integrated with Insider’s systems through the Model Context Protocol (MCP), IAra began accessing information such as order history, inventory, and customer preferences. |
“The biggest mistake is thinking that AI is only for customer service. Good AI is the one that speaks before the customer complains.”
Cacau Pellegrino
Head of CX at Insider
Transformation into numbers
The results show that automation was not just a technological experiment, but a structural change in the way Insider's support operates. With the evolution of the ClaudIA phases and the creation of IAra, the operation began to resolve a large part of interactions directly through AI, maintaining high levels of satisfaction and freeing the human team to work on more strategic fronts.
Metric | Result | Impact |
|---|---|---|
Support cases resolved by AI | 76% | Freeing the human team for more complex cases. |
CSAT in AI interactions | 81% | High satisfaction even in automated support. |
Conversion with IAra (A/B tests) | +18% | Support began to directly influence sales. |
CSAT response rate | 27% | High customer engagement in evaluations. |
With AI taking on a significant portion of interaction volume, the team began to dedicate more time to customer relationships, pattern analysis, and journey improvements. This brought more predictability to operations and allowed the company to scale support without putting pressure on the team structure.
The combination between AI and human operations transformed Insider's CX into a structure prepared for accelerated growth. Today, the company can absorb significant volume increases without compromising the customer experience or needing to expand the team in the same proportion.
“Today we scale without fear. The operation can handle 3x growth without needing to double the team.”
Cacau Pellegrino
Head of CX at Insider
What Insider learned by implementing AI in customer service
Topic | Description |
|---|---|
AI doesn't fix bad processes | Automating disorganized customer service doesn't solve the problem; it only scales it. Before expanding the use of AI, Insider needed to better structure processes, flows, and service criteria. |
CX is a strategic asset, not just an operating cost | By analyzing data from customer interactions, the team began identifying opportunities to improve product, communication, and the purchase journey. |
AI curation is an ongoing process | Artificial intelligence is not a project with a beginning, middle, and end. The assistant's performance depends on constant review of responses, training, and monitoring of interactions. |
The role of the CX leader is changing | The operation now requires professionals capable of combining customer service management with data analysis, process understanding, and the use of AI tools. |
“The secret isn't in the AI, it's in how you use it. ClaudIA (or IAra) reflects our culture. And culture is what ensures that the customer continues to love Insider, even when service is handled by a machine.”
Cacau Pellegrino
Head of CX at Insider
Cloud Humans answers
Did AI replace Insider's customer service team?
No. AI began to act as the first layer of support, handling recurring and operational questions. The human team began focusing on more complex interactions, customer relationships, and improving the experience.
How was AI trained to serve Insider's customers?
AI was trained using the operation's service history, combining real interaction data with rules and curation from the CX team
Was AI implementation immediate?
No. The evolution happened gradually, through different stages of maturity, starting with resolving recurring questions and later supporting the purchase journey within e-commerce.
How does Insider ensure the quality of AI responses?
The team maintains an ongoing process of curation and review of interactions. This monitoring makes it possible to identify new customer questions, adjust responses, and ensure that the assistant's tone remains aligned with the company's human customer service.
