Elysia, AI emotional support for women
Women's wellness platform with AI, founded, designed, built and shipped. From problem to product in production.
A personal problem turned into a company
Elysia was the first product of OutStress, a company I founded. It started from a simple observation: women around us who were clearly struggling emotionally, but who never asked for help. Not from friends, not from family, not from professionals.
The gap wasn't just about access to therapy. It was about something harder to solve: the willingness to admit they needed help at all.
30 conversations. One insight that changed the product.
Before designing anything, I spoke with approximately 30 women in the Torres Novas region. The goal was to understand how they dealt with emotional difficulty and what prevented them from seeking support.
I expected to find practical barriers: cost, waiting lists, distance. What I found was more fundamental.
The majority of women interviewed said they habitually ignored their own feelings. They described putting on a mask, not wanting to worry others, and feeling that their problems weren't serious enough to justify professional help.
This reframed the design problem entirely. The challenge wasn't to make therapy more accessible, it was to create a space where a woman could first admit, even to herself, that she was struggling. Without shame, without judgment, without needing to explain herself to another human being.
Every decision followed from the research insight
AI first, human second
The AI assistant was the primary interface by design, not a fallback. Women who couldn't speak openly to a person could speak to something that would never judge and was always available. A psychologist would only enter the flow when the AI detected signals requiring professional intervention.
Soft language, not clinical language
Every word in the interface was reviewed to remove clinical framing. 'How are you feeling?' instead of 'Describe your symptoms.' Warmth was a functional requirement, not a stylistic preference.
Gradual onboarding, zero pressure
The onboarding never asked for commitments. No goals, no schedules, no health assessments upfront. The app earned trust incrementally, the first interaction was designed to feel like a gentle check-in, not a triage form.
Holistic platform: AI + science + community
The product integrated chronobiology and crononutrition to personalise content to each user's biological rhythms. We added online therapeutic circles, a cyclical journey feature, and exclusive content, building a complete platform, not just a chatbot.
Everything, from concept to product in production
As founder and sole designer of OutStress, I owned the entire product surface: user research, product strategy, UX flows, visual design, interactive prototyping, brand identity, and development of the application for both iOS and Android.
There was no team. Every decision was mine to make and mine to defend, to the women we interviewed, to the incubation programme, and to the market when we launched.
Launched. Validated. Discontinued.
30+
Women interviewed in primary research
iOS + Android
Launched on both stores
Incubated
Startup Torres Novas programme
Full stack
Research to code to production
The product launched on both stores and was validated by the market. OutStress was accepted into the Startup Torres Novas incubation programme. The app was discontinued due to lack of financial resources to scale.
What I carry from Elysia: the hardest product problems aren't about features. They're about the emotional conditions under which someone will even try to use what you built. And that it's possible to go from zero to shipped product alone, as long as the problem is real.