runtime

A centralized place for digital volunteering tracking

Comment

Inspiration

After working on a volunteering project for an organization focused on tech. We realised that it was extremely tedious to keep track and log all the apps that were used. What was even more difficult with keeping track was workflows that used more than one app. All of this culminated into one idea at this hackathon, find a solution to make keeping track of volunteering hours that are completed digitally easily and credibly.

What it does

Runtime provides a centralized platform for coding volunteers to track their volunteering hours and supervisors to keep track of their volunteers. If you are a volunteer, . If you are a supervisor, log in to the runtime dashboard web app to view your volunteers' data, see analytics, approve shifts, run AI assessments, and generate AI weekly digests.

How we built it

We mainly used AI coding assistants like Claude or Codex to build out the foundations for our project. We then bug-tested the issues and sent them back to our assistants to help debug.

Challenges we ran into

While building this, we ran into several challenges with the databases and APIs. Focusing first on APIs, when we were testing out our Gemini API function in our software, we found that it wouldn't work giving us a 501 error where it told us that there was high usage and we couldn't use the API. We changed models, looked through menus in Google AI Studio, but couldn't find a solution until a mentor helped us out. The solution was simpler than we thought, we just had to use a worse model (3.8 flash -> 3.5 flash lite). For our desktop agent, we found that it wouldn't show the total time worked and the time worked in a single day. We tested if the issue was with the database being setup correctly but it was sending out correct data. The main issue that caused all of this was a variable in the agent that didn't pass the user id. This causes the endpoint to return a 422 which turns into a 0. The fix was to make the user id optional and make it default to the authenticated user.

Accomplishments that we're proud of

We are proud of the fact that we actually have a working backend for this project. Setting up the backend was pretty difficult for us, but with the help of Claude Code and other AI models, we were able to. In addition, we are proud that the front-end of the dashboard looks minimalistic and modern.

What we learned

We learned how to set up and run Tiger Data and Gemini API for backend processes.

What's next for runtime

If we perfect runtime, we could possibly release it to the public. One improvement we could make to make it look more professional is to make the database have its own site instead of localhost.

Built With

Share this project:

Updates

Submission history