AI-Powered KosmoHealth Platform: Revolutionizing reproductive health support with emotion recognition and multilingual chatbot for Rwanda

Many women and girls in Rwanda, particularly in rural areas, lack access to reliable reproductive health information, affordable menstrual products, and medical support. This leads to period poverty, misinformation, and preventable health issues. There is a strong need for inclusive and sustainable solutions to address menstrual hygiene, reproductive health, and maternal care.

• Enhance Kosmohealth platform with AI for reproductive health guidance. | Milestone: Launch the fully functional AI-enhanced version of KosmoHealth by Q3 2025, supporting English and Kinyarwanda.
• Strengthen partnerships with health professionals and organizations. | Milestone: Onboard 50 certified health professionals and Institutional partners by 2026.
• Expand access to reproductive health support in Rwanda through KosmoHealth mobile platform and hotline. | Milestone: Reach 10,000 users in Kigali by the end of 2026.

With this project we aim to have:

1. AI-Powered Emotion Recognition and Face Detection: We use TensorFlow.js and face-api.js to detect faces and recognize emotions like happiness, sadness, anger, or anxiety during live video consultations.

This helps health professionals respond more empathetically during sensitive topics such as menstruation or maternal health.

2. Automatic Privacy-Focused Face Blurring: If discomfort is detected via emotion cues, the system automatically applies a blur effect on the user’s video feed.

This is processed locally in the browser for full privacy and no external data sharing.

3. AI Chatbot in Kinyarwanda and English: We’ve integrated a multilingual AI chatbot that can assist users in Kinyarwanda and English.

The chatbot helps with:

• Answering basic health questions.
• Providing menstrual and maternal health information.
• Guiding users through app features and support services.

This feature ensures inclusivity and accessibility for local communities who may not speak English fluently.

4. Emotion-Triggered Support Prompts: The system can trigger alerts or support options for healthcare workers when emotional distress is detected, enhancing real-time care and response.

Health Professionals: 50+ doctors, nurses, and counselors providing consultations and support via the platform.

Women and Girls: 10,000+ users in Rwanda seeking reproductive health guidance and information.

Partner Organizations: 10+ organizations promoting KosmoHealth in rural communities.

Support Team: 5+ members managing system operations and troubleshooting.

AI chabot Integration Mobile App Development User Privacy Protection user support training Partnerships Development promotion campaign Existing Feature updates App maintainance Data sourcing in kinyarwanda for more training.

Internal Impact: Improved user experience: AI-enhanced emotion recognition and local face blurring make consultations more sensitive, private, and human-centric.

Efficiency in care delivery: Health professionals are better equipped to respond empathetically and promptly through emotion-triggered prompts.

Innovation edge: Local processing ensures high data privacy compliance, boosting user trust and positioning the platform as a tech-forward solution in digital health.

Bilingual chatbot reduces load on staff by resolving routine queries automatically and at scale.

External Impact: 10,000 new beneficiaries reached in Kigali and rural communities via mobile platform and hotline.

Inclusive support for non-English speakers through AI chatbot in Kinyarwanda.

Enhanced reproductive health literacy and menstrual hygiene awareness, especially among youth and underserved populations.

Faster access to health professionals with real-time, personalized guidance and follow-up support.

Better maternal outcomes through targeted information and emotional care, especially for first-time or at-risk mothers.

Hosting and domain Content (images, text) Data set(s)