On April 13-14, Milena Miernik, who is doing her LiA at Aixia, participated in the HackHERHealth hackathon event, which had the overall goal of reducing the gender gap in medical health. Women are underrepresented in medical studies resulting in less support and less effective treatment for them. Over 200 participants gathered both on site at Sahlgrenska Science Park and online to jointly address this problem through the use of AI solutions.
The hackathon was divided into five different challenges, each focusing on different aspects of women’s health and well-being. I contributed to Challenge 2, where my team developed a solution that supports neurodivergent girls in their everyday activities. AI assistant named “Jackie” that helps with clothing selection and daily activities, reducing user decision fatigue and cognitive overload.
The use of AI in this app has two main advantages: the ability to analyze and take into account nearly 30 different weather parameters to provide forecasts of appropriate attire and the flexibility over time as the user provides their own input and the model can be adapted and trained accordingly.
Integration with Azure tools (including Cognitive Search and AI Speech) also makes it possible to communicate with the app about the weather, as well as receive recommendations for location- and weather-appropriate activities to improve well-being. The RAG approach to chat makes this app very secure, as responses are always based on curated sources.
The image below shows the rough idea of the back-end AI architecture, but if you want more details, don’t hesitate to contact me. I am happy to explain more!
