Inspiration
- We wanted to find a better a way to explore the variety of living things around us in a fun and interactive way!
What it does
- BioBlender takes any two user inputted animals and returns an informational page on an animal with traits similar to both of the inputted animals.
- The information gathering is sourced from the OpenAi-API.
How we built it
- This app was developed using primarily Android-Studio
- General Flow: Take in user animal inputs (or random animal) → Make an OpenAi-API call to determine unique traits pertaining to each animal → Comparison algorithm is ran to create an API call prompt that returns a unique animal each time → OpenAi-API call is made to discover interesting facts pertaining to the newly discovered animal → Make a SerpApi call to receive a random popular image URL of 'X' animal from google → Display all animal information onto the final information page.
Challenges we ran into
- Difficulties running OpenAi API on android studio
- Challenges with merging file information from different members using Git
- Small bugs and errors hindering programming progress (ie: providing android studio with permission to use the internet in order to make API calls locally)
- Being able to randomly source an image for each animal (URL's varied greatly and all images were loaded instantaneously as opposed to pre-launch)
- Finding metrics/information that can be used to repeatably and reliably compare animals
Accomplishments that we're proud of
- We set realistic goals and managed to complete them within the constrained timeline of the event, through planning and determination
- We came up with an interesting idea that we considered unique and insightful
What we learned
- We all learned how to use and became proficient with android-studio with very little prior experience
- We learned how to install and make APIcalls
- We came across many new animal and learned about many of their unique attributes during the development process
What's next for BioBlender
- Game Mode (combining a list of randomly generated animals with the goal of reaching an imposed destination animal), Game Mode could also include an achievement system which would give the user in-app rewards for meeting certain criteria with their pool of generated animals (eg: 10 animals from South America, or 5 animals that have sharp claws).
- Save feature to save the generated information page for unique and interesting animals.
Built With
- aiohttp
- android-studio
- chaquopy
- java
- openai-api
- python
- serpapi

Log in or sign up for Devpost to join the conversation.