Inspiration

In Liberia, where 32% of adolescents face unintended pregnancies and SRHR information remains scarce due to stigma and limited healthcare access, we saw an urgent need for discreet, accessible solutions. Many young Liberians travel hours clinics only to find them closed or lacking privacy. Our team, having witnessed friends and family members struggle with SRHR challenges, was driven to create a solution that puts dignity, privacy and accessibility at the forefront.

What it does

Konektoh is a comprehensive mobile platform that provides:

AI-Powerd SRHR Assistant: 24/7 confidential chatbot providing culturally-sensitive SRHR information in simple Liberian English. Offline-First Clinic Finder: GPS-enabled map showing One-Stop Centers with routing, even without internet. Health Tracker: Private symptom and period tracking with educational insights. Emergency Hotlines: Immediate access to verified SRHR support service. Hybrid Architecture: Seamlessly switches between online AI and offline knowledge base.

How we built it

Our technical stack combines cutting-edge AI with robust mobile engineering: ** Frontend Architecture:

React Native/Expo with TypeScript for cross-platform compatibility. Custom Navigation with React Navigation for smooth user experience. Map Integration using React Native Maps with OSRM for free routing. Offline Storage with AsyncStorage and SQLite for data persistence.

Backend & AI Infracstrure

Flask microservice with Ollama integration

AI/ML Pipeline:

Ollama with Phi model for local AI inference. Dockerized services for easy deployment. Hybrid Response System combining online LLM with offline semantic search. TensorFlow.js for future on-device model deployment.

Challenges we ran into

Technical Challenges:

Connectivity Reality: Liberia's internet penetration is ~30%, requiring sophisticated offline capabilities. Model Optimization: Balancing AI accuracy with device resource constraints. Geospatial Challenges: Accurate clinic mapping in areas with limited digital infrastructure. Cross-Platform Consistency: Ensuring uniform experience across diverse Android devices.

Cultural & implementation Challenges:

Language Sensitivity: Translating complex medical terms into accessible Liberian English. Trust Building: Designing interfaces that overcome stigma and encourage engagement. Data Privacy: Implementing robust security in resource-constrained environments. Community Integration: Ensuring solutions align with local healthcare ecosystems.

Accomplishments that we're proud of

*Functional Hybrid AI System delivering 95% uptime regardless of connectivity. *Cultural Competence validated by preliminary user testing with Liberian youuth. *Complete Technical Stack from AI backend to polished mobile interface. *Real Navigation using free OSRM services, eliminating API costs. *Privacy by Design with zero sensitive data transmission without consent. *Scalable Architecture ready for deployment across West Africa.

What we learned

Technical Insights:

Offline AI requires careful knowledge base curation and semantic matching. Docker containerization dramatically simplifies AI model deployment. React native's performance limitations require creative optimization for maps.

Cultural Learnings:

Trust is built through languages simplicity and visual familiarity. Community health workers are crucial for adoption and validation. Privacy concerns often outweigh feature desires in SRHR contexts.

Imapct Realizations:

Digitalizations solutions must complement, not replace, human touchpoints. Sustainability requires partnership with existing healthcare infrastructure. Youth engagement thrive on non-judgemental, accessible design.

What's next for Konektoh - Your Digital SRHR Companion for Liberia

Short-term (3-6 months):

Pilot deployment in 3 Liberian counties with YMCA partners Integration with national SRHR hotline services. Expanded offline knowledge base covering 500+ SRHR topics. Medium-term (6-12 months):

SMS-based access for feature phone users. Healthcare provider dashboard for follow-up care. Multi-language support (French, local dialects). Long-term Vision:

Expansion to Sierra Leaone and Guinea. Predictive analytics for public health planning. Blockchain-based anonymous health records. Integration with national health management systems.

Click on the link below to install the android app, and experience something real

Built With

Share this project:

Updates