Inspiration
Our vision was simple: to create a platform where people can effortlessly generate old newspaper articles based on both past world events and personal milestones causing a nostalgic feeling.
What it does
When the user first logs in, they will be creating a secure account with auth0. This will give them the assurity that the photos that they upload will be secure. Once they upload every photo they want to incorporate into their photo album, they can then proceed to the album tab and generate a nostalgic newspaper of personal photos and world events for that specific year.
How we built it
The user first logs in and create a secure profile on our app, using auth0. When they create their profile, they will upload all the photos they want to include in their photo album. These photos when uploaded will automatically be summarized using GPT-4. The images are subsequently saved in AWS s3 for usage later on. The text summary and the the url to the images in the s3 bucket are saved in firebase.
When the user wants to create a photo album for a specific year, they choose their desired year and the cohere AI model searches the web for major events from that year and combines their personal events along with world events into one nostalgic photo album that the user can download and keep.
Challenges we ran into
We ran into several challenges along the way, and every single obstacle represented a unique learning opportunity for us. Each team member learned something new and levelled up their skills. Whether it be learning Flask and React, or struggling to integrate these two frameworks, we faced it all. However, understanding Flask and how it works was one of our bigger challenges for Fullstack development.
Accomplishments that we're proud of, and what did we learn?
For the first time cohere, auth0 and gpt-4 were used in development and a single app was created using these frameworks. We were able to create 2 separate instances of our website, one with Flask and one with React. These two are yet to be integrated, however, we are proud that we were able to leverage each individual frameworks strengths and use them.
Creating the AI backend with python was really cool and a great learning opportunity as this was a first for our team member.
What's next for Blast from the Past
The next steps would be finishing this app and creating a fully functional nostalgic photo album. Implementing the upload photo button, download album button and actually creating a working database with firebase would be our top priorities.
Built With
- auth0
- cohere
- css
- figma
- flask
- gpt4
- html
- javascript
- python
- react
- typescript
Log in or sign up for Devpost to join the conversation.