SUCCESS STORY
The company used AI to automate recurring questions, reduce first response time, and free up the support team for more complex and strategic support requests. See the results achieved in just 8 months!
Metric | Result |
|---|---|
Cases resolved by AI | 60% |
Automated cases | 32,778 |
Nina CSAT (AI agent) | 72% |
About Kenlo
Kenlo is a technology company that develops solutions for the real estate market. Its purpose is to simplify and digitize the journey of buying, selling, and renting properties throughout Brazil. Since the real estate market has a highly consultative profile, support has always played a central role in the company's operations. “Brokers and real estate agencies frequently turn to the team to clarify questions about processes, settings, and use of the platform,” said Tatiana Dias, Support Supervisor at Kenlo
Reactive mode: growth versus operations
The operation dealt with a high-volume, low-predictability scenario. Internally, the team described the routine as a true “firefighting mode,” because the high volume of requests created constant queues and made it difficult to prioritize more complex cases.
+6 hours of first response time
Even for simple questions, customers waited hours.
Constant reactive mode
Firefighting team. No time for strategic actions, always putting out fires.
+50% repetitive volume
More than half of the tickets were questions and guidance.
Low CSAT
Compromised agility affected the customer experience.
The first attempts to scale support
Faced with this scenario, Kenlo began looking for alternatives to reduce wait times and ease the pressure on the support team. But the problem was not just the volume of tickets, but their nature.
More than 50% of the volume consisted of simple questions and guidance that consumed the team's time and made it difficult to prioritize more complex cases.
Step | Title | Description |
|---|---|---|
01 | Dedicated triage team | Iara was simple: filter basic operational questions and route only the most complex cases to the main support team. This approach helped reduce part of the queue, but it did not solve the structural problem. |
03 | Conventional bots | For the bot to work properly, the customer had to type exact terms. So it was necessary to register several variations of the same question, which made maintenance complex and inefficient for the volume of tickets the company received. |
Even with these initiatives, the team still faced a recurring challenge: balancing customer base growth without having to increase the team proportionally. To do that, it was necessary to find a solution that could truly absorb the volume of operational questions without compromising the customer experience.
Kenlo + Cloud Humans: a strategic partnership
It was in this context that Cloud Humans joined the operation, helping to completely transform the way Kenlo serves its customers.
Key gains:
First response time dropped from 6+ hours to 20 seconds.
+60% of tickets resolved directly by AI.
AI came to be seen as a partner, not a replacement.
CSAT steadily increasing (70% → 72%).
Faster onboarding of new employees.
Freed up the N2 team to focus on critical support and customer relationships.
“Cloud Humans' entry was truly a turning point for our operation. It was a two-way process: they brought all the technical expertise, and we added deep knowledge about our customers.”
Tatiana Dias
Support Supervisor at Kenlo
The 4 phases of transformation
The 4 phases of transformation
The implementation of Cloud Humans' AI at Kenlo was carried out gradually, combining technological adaptation and the operational reorganization of the support team.
Preparation & Context
Importing history and defining Nina's tone of voice (Kenlo's customer service AI).Setup & Customization
Customization for a consultative audience and older audience.
Testing & Adjustments
Integrated auditing and continuous feedback for improvement.
Go Live & Evolution
Launch with a dedicated cell for continuous auditing.
Fun fact:
Internally, the assistant came to be called Nina. The name and identity of the AI were defined through a collaborative process within the company, involving different areas in the choice of the avatar and the assistant's personality. The initiative helped bring the technology closer to the team and reinforced the idea that the AI would be one more member of the operation.
The evolution of the operation also led to the creation of a role dedicated to auditing and continuous improvement of the AI. This role became responsible for reviewing interactions, adjusting content, and ensuring that the assistant's tone of voice remained aligned with Kenlo's human support.
“When many people train the AI at the same time, each one does it differently. We realized we needed to centralize this to maintain the same service standard.”
Geovanna Xavier
Support Analyst responsible for this area at Kenlo
Transformation into numbers
In a short time, Kenlo's support operation underwent a significant change in how service requests were distributed between humans and automation. The numbers below show some of the results achieved in 8 months:
Metric | Before | After | Improvement |
First response time | +6 hours | 20 seconds | 1,080x |
Tickets resolved without escalation | ~0% | +60% | +60% |
Total retained tickets (Jan-Aug 2025) | 0 | 32,778 | Release |
CSAT (Customer Satisfaction) | ~60-65% | 72% | +12% |
Team operational workload | 100% manual | 20% manual | -80% |
AI began to absorb a large share of the simpler operational questions, drastically reducing response time and freeing the human team to handle more complex analyses and interactions that require context and relationship-building
with the customer.
The result was a combination of operational gains and improved experience: faster responses for the customer and more strategic focus for the support team.
What changed in the team's mindset
Automation = Freedom, not reduction
AI doesn’t replace people. Those who fall behind are those who don’t learn to work with it. Now the team can focus on critical analyses, internal improvements, and more complex cases.
Early involvement
When the team feels ownership of the project, adoption is natural and success comes much faster.
CX with data is strategy
With the reduction in operational volume, the team began analyzing patterns in questions and identifying opportunities to improve the product.
Tone of voice matters
Clear communication, without technical jargon. Our audience is consultative and needs simplicity.
“If I could leave one piece of advice for other leaders, it would be: don’t see AI as a replacement, but as a partner. The big turning point happens when the customer realizes they gained agility without losing the human warmth of service.”
Tatiana Dias
Support Supervisor at Kenlo
Cloud Humans responds
Did AI replace Kenlo’s support team?
No. AI started handling repetitive and operational questions, allowing the human team to focus on more complex cases, critical analyses, and customer relationships.
How long did it take to implement AI?
The initial implementation took about five weeks and was divided into four stages: preparation, customization, testing, and launch.
What was Kenlo’s main gain with AI?
The main gains were: 1) a reduction in response time from over 6 hours to about 20 seconds. 2) more than 60% of tickets automatically resolved. 3) more than 32,000 support interactions retained in 8 months.
Does AI work for less digitally savvy audiences?
Yes. AI was trained with a simple, consultative tone, suited to the traditional profile of the real estate market.
