Nkwa Hia

Inspiration

Nkwa Hia means life is precious. We built a web app to synchronize information between hospitals, EMTs, and patients. Our inspiration came from a March 23, 2026, press statement by emergency medicine residents in Accra. They reported patients receiving care on floors at Korle-Bu Teaching Hospital. They described a fractured national response system where primary facilities dump patients at tertiary centers. Where pre-hospital coordination is absent and no national bed-tracking exists, we built Nkwa Hia to fix the information gap.

What it does

Our platform provides a real-time dashboard for the Ghanaian healthcare grid. We created an AI triage service that analyzes a casualty's symptoms. It recommends a hospital based on severity, location, and drug availability. You get a precise destination, without wasting time at facilities without oxygen or surgeons. A live heat map shows hospital bed availability across the city, showing EMTs and 193-callers see exactly where resources are located.

How we built it

We built the front end using React for a fast, responsive user interface. Our back end uses a Python framework to manage the data flow between facilities. The system can switch between Gemini and Claude AI models to process symptom reports. We used a three-tier architecture to ensure reliability and security of sensitive data. We mapped several public facilities in Accra into a single database.

Challenges we ran into

We are realistic about the limitations of AI. Even Amodei notes that intelligence is not magic fairy dust. It hits physical limits. An AI can recommend a hospital, but it cannot create a physical bed or a doctor. We faced data quality issues during testing. If a hospital fails to update its inventory, the AI makes a bad choice. We also resourcefully introduced a switching system between AI models, to account for the limitations of any one particular model. We further integrated a manual override for clinicians to ensure human judgment remains the final authority.

Accomplishments that we're proud of

We successfully bridged the communication gap between EMTs and ER staff. Our system allows hospitals to prepare for critical arrivals before the ambulance pulls in. We pushed ourselves beyond our comfort zones learning full-stack web development on a tight schedule. We believe our project reflects the love and work we poured into it.

What we learned

We learned that systemic reform requires more than code. It requires a commitment to truth. Some hospital managements prefer "headline figures" over transparent data. We learned that AI must assist, not replace, medical experts. Ethics require transparency. You see why the AI chose a facility, but you also verify the data. We also learned that local maintenance is vital for long-term success. Technology fails without people who believe in the mission.

What's next for Nkwa Hia

We plan to expand our heat map to include private clinics and specialized labs. We want to integrate wearable health data into the 193 caller profile. This will provide the AI with pre-existing condition data for better triage. Amodei predicts AI will improve governance and peace. We start with the ER. We aim to scale Nkwa Hia to other regions in Africa and the world. We believe life is truly precious, and we want the system to reflect what matters.

Built With

• antigravity
• bycryptjs
• claude
• cors
• dotenv
• fastapi
• firebase
• firestore
• gemini
• github
• google-cloud-run
• google-maps
• javascript
• lucide
• python
• react
• typescript
• uvicorn
• venv
• vite