VolunTold

A platform that helps make a difference, anywhere.

Comment

Inspiration

Former Civic Hacking Projects: We found inspiration from former projects that helped us make a difference it helped us realize how we can make a impact that matters.

Lack of Volunteer Apps: We couldn't think of a good way to find a volunteer opportunities in our community, so we decided to make one.

Lots Of Redbull: What can we say? We needed the ENERGY

What it does

VolunTold has several features based on making and fostering a community of VolunTold's users.

Gathering Volunteer Opportunities: In the Discover tab you can select volunteer opportunities gathered by Google Gemini based on your location. This can work with any link, so we added a feature to get the community involved and share their local volunteer places.

Creating Volunteer Opportunities: In the tab Opportunities, you can select volunteer opportunities that are exclusive to the web app. This allows organizations to post volunteer opportunities and in the future push users with important notifications for a important volunteer work needed after a crisis.

How we built it

Layer Technology Description
Frontend 🧱 Bootstrap CSS framework for responsive, mobile-first UI design.
🧩 Flask Templates (Jinja2) Server-side rendering and dynamic HTML generation.
Backend 🐍 Flask Lightweight Python web framework handling routes and APIs.
🗃️ SQLAlchemy ORM providing structured access to the database.
💾 SQLite Simple, file-based database for local persistence.
Infrastructure 🐳 Docker Containerization for consistent and portable app environments.
☁️ Vultr VM Cloud host running the Docker image with Flask and SQLite inside.
AI Integration 🤖 Google Gemini with Open Router Powers the Volunteer Gatherer — AI for data parsing and matching.

Challenges we ran into

Framework Learning Curve: It was our first time using these web frameworks to create an application

Environment Inconsistencies: Local developer environments were somewhat inconsistent, especially with regards to database schemas at different points in development

Webscraping Limitations: OpenAI API was misbheaving, and data retrieved was sometimes inconsistent, so we had to refine our original prompt and scope a bit in order for the webscraping module to work

Accomplishments that we're proud of

AI Discover Feature for creating summaries of possible volunteer opportunities given a website

Dockerization for consistent builds and local developer environments (via docker-compose)

Framework zero to hero over the course of the hackathon

What we learned

LLM Interactions using available web APIs

Flask backend for quick and easy API's / webservers

Cloudflare for domain resolution and secure site connections

What's next for VolunTold

Try our docker-compose:

services:
  voluntold:
    image: ghcr.io/savern02/hackokstate:latest
    ports:
      - "5000:5000"
    volumes:
      - ./instance:/app/instance # <-- you will need an sqlite database
    environment:
      OPEN_AI_KEY: "" # <-- fill with an OpenAPI key for the webscraping summaries

Note: certificates for demo site set to expire a few days following the competition

Share this project:

Updates