One-Stop-Story

An AI-powered app that helps homeless individuals get aid faster, it listens once, fills forms, finds nearby shelters, and connects them directly. “Tell your story once, get help instantly.”

Comment

Inspiration

The project was inspired by the real struggles faced by people experiencing homelessness or crisis. Many individuals are forced to retell their stories multiple times to different agencies just to access basic services like food, shelter, or healthcare. We wanted to build a compassionate AI solution that listens once, understands, and connects users to the right help, fast, respectfully, and securely.

What it does

Our app allows users to speak or type their situation once, and the AI automatically: Transcribes and structures their story. Fills out necessary intake forms for multiple agencies. Finds matching shelters, food centers, or medical clinics nearby. Let users call or email facilities directly with a pre-filled PDF of their details. For social workers and agencies, it streamlines case management, generates consistent records, and speeds up service delivery, helping communities respond with empathy and efficiency.

How we built it

We used a full-stack AI-driven architecture combining: Frontend: React Native + Expo for a cross-platform mobile app. Backend: Node.js + Express handling API routing, AI logic, and PDF automation. AI: Google Cloud (Vertex AI + Speech-to-Text) for transcription, translation, and NLP-based data mapping. Document AI: Bild.AI for OCR, validation, and automatic PDF generation.(Future Enhancement) PDF Engine: pdf-lib for dynamic form creation and export. Security: HTTPS, .env secrets, and field-level data consent. Mathematically, structured mapping uses a classification approach to extract entities like: Needs = {shelter, food, safety}, Risks = {violence, medical} Needs={shelter, food, safety},Risks={violence, medical}

Challenges we ran into

Accurately interpreting unstructured or emotional speech into actionable data. Maintaining data privacy and user trust across multiple agencies. Generating consistent AI-filled PDFs from diverse templates. Integrating multiple APIs (Google Cloud, Vertex.AI) within tight security constraints.

Accomplishments that we're proud of

Developed a working prototype that turns a person’s spoken story into a structured case file. Created end-to-end automation from speech → structured data → pre-filled PDF → direct contact. Demonstrated cross-agency interoperability for social services. Built a solution that brings technology and empathy together to support vulnerable communities.

What we learned

We learned the importance of human-centered AI design, how a well-designed workflow can restore dignity to those seeking help. Technically, we deepened our skills in NLP mapping, PDF automation, cross-platform mobile development, and secure AI integration.

What's next for One-Stop-Story

Integrate Stripe APIs for verification of NGOs. Integrate geo-location APIs for real-time nearby service suggestions. Add push notifications for appointment reminders. Expand OCR and translation capabilities using Bild.AI. Move to Firebase/MongoDB for scalable storage. Develop an analytics dashboard to help agencies understand needs and allocate resources more effectively.

Built With

  • code
  • environment-variables-(.env)-other-tools:-axios
  • expo-av
  • filesystem-security:-https
  • languages:-javascript
  • qr
  • react-native-linking-api
  • speech-to-text-document-intelligence:-bild.ai-pdf-tools:-pdf-lib
  • typescript-frontend:-react-native-+-expo-backend:-node.js-+-express-ai-&-cloud:-google-cloud-vertex-ai
Share this project:

Updates