Overview
MenteSana is a web platform aimed at teenagers that seeks to promote emotional wellbeing, reduce stigma around mental health and facilitate access to reliable resources. It allows users to register daily how they feel and receive personalized responses generated by artificial intelligence, as well as access information, events and educational content about mental health. The project was funded by the Eureka program, validating both its social impact and its technological approach.
Problem
Many teenagers lack safe spaces to express how they feel; they avoid talking about mental health due to stigma or fear of judgment; they don't easily access reliable resources or early support. Existing solutions tend to be clinical, less accessible or not adapted to adolescent language and digital habits.
Solution
We designed a platform where users register their emotional state daily; receive personalized messages generated by AI based on their state; access clear information about mental health; find local events and resources; interact anonymously and frictionlessly.
Process
- 1Research and MVP definition focused on adolescent usability
- 2Design of onboarding flows and emotional logging (Sofía)
- 3OpenAI integration and persistence backend (me)
- 4Joint development of landing and key modules
- 5Deployment and institutional validation through Eureka funding
Results
Project in production. Funding obtained from the Eureka program. Active platform with real usage. Validation of AI applied to emotional wellbeing.
My Role
I was responsible for: backend architecture and data model; OpenAI integration for empathetic personalized response generation; API development and business logic; implementation of key landing modules; deployment and infrastructure. Collaboration: project developed with Sofía Jamanca (Frontend Engineer), who led UX/UI design, complete frontend development, onboarding flows and user experience.
Technical Architecture
Frontend: React · Vite · Tailwind CSS Backend: Node.js · Express · PostgreSQL AI: OpenAI API for personalized message generation based on emotional state
Major Challenge
Design of emotional responses with responsible AI: I designed prompts, rules and validations so that generated messages are empathetic, non-pathologizing, appropriate for teenagers and do not replace professional support. Secure backend for sensitive data: I implemented secure persistence, separation of personal data and flows that prioritize privacy and anonymity.
Key Learnings
- Collaborative work on product
- Responsible integration of AI in sensitive contexts
- Backend design for social impact applications
- Building MVPs with institutional validation