Nilo Chat
The interface between patients and their care team - redesigned for the age of AI.
nilo
2022 - Redesign
Overview
The Chat tool is currently the most used platform at Nilo. However, it had many usability and interactivity problems.
Some professionals did not feel confident using it and also felt that it lacked many basic features for a chat tool.
My Contribution
My goal in this project was to solve old version Nilo Chat's usability problems.
To achieve this I did several interviews with users and usability tests that served as the basis for the flow designs and visual updates of the product.
My Role
Product Designer
Duration
4 months
Year
2022
Tools
Figma
nilo-chat-nurse
What they say
- "Sometimes I need to know who's been waiting the longest so I can prioritize service"
- "I need to know how much time is left to close the patient contact window"
- "I would like to see inbox and opened conversations on the same screen"
nilo-chat-ui-overview
Old version
nilo-chat-oldversion
nilo-chat-feature1
Gets care team closer to the patients
nilo-chat-notes
Take notes and share with colleagues
nilo-chat-timer
Never loose the open window to start a conversation with a patient
nilo-chat-priority
Mark some patients as priority, so they can be attended first
nilo-chat-audio
Send and receive audio and video messages
nilo-chat-waiting
Get notified about who is waiting longer
Where it all started
With usability solved, we had a solid foundation. When AI capabilities emerged, we saw an opportunity to take the platform further.
2025 - Agentic Mode
Overview
The existing chat interface was not designed to support agent interactions, handoffs, or guardrails.
We needed to redesign the entire experience, making it intuitive for patients and easy to maintain for the product team - without compromising clarity, empathy, or efficiency.
My Contribution
As the product designer, I led the redesign, collaborating closely with engineering and our conversational AI specialist. I worked on:
- 路 Mapping agent capabilities and defining use cases for triage and scheduling.
- 路 Designing an adaptive UI that could transition smoothly between agent and human interactions.
- 路 Ensuring guardrails, fallbacks, and agent transparency (e.g., showing who is responding).
- 路 Supporting our new design system (MUI) to ensure consistency and scalability.
My Role
Product Designer
Duration
4 months
Year
2025
Tools
Figma, ChatGPT
Impact
The new experience was piloted with Fleury Group, a major healthcare provider in Brazil. After 4 months of usage, the AI agent reduced human triage and scheduling requests by 96%, freeing up resources and enabling 24/7 patient support.
This redesign became the foundation for future AI initiatives, aligning product design with the company's long-term automation strategy.
Media feature:
How Nilo is Connecting 3.5 Million Patients Using AI
nilo-chat-media
Interface overview
nilo-chat-interface
Exception Flows and Contingency Scenarios for the Virtual Agent
Mapping key support contexts to ensure appropriate agent behavior in cases like AI latency, system failures, human fallback, and safe conversation recovery.
nilo-chat-exceptions
More productivity and organization
Everything they need to know to make the care experience more efficient, in a beautiful and organized screen.
nilo-chat-final