Inspiration
We found Corgi(our puppet unicorn) discovering Berlin. And we thought that we should build something that would helps us to find more unicorns.
What it does
We do voice recognition in a web app that detect continuously words from the startup and tech world (and maybe some buzzwords) and at the end of the pitch completes a ranking that can be compared to the other pitches/presentations.
How we built it
We set up a web app, cross functional along several devices web compatible. We use webkit speech recognition, we implemented a nlp pipeline and we use D3 for the data visualization. We use local storage for the data persistence.
Challenges we ran into
Doing all the work in the frontend. Hacking voice recognition to work properly without an external provider.
Accomplishments that we're proud of
Building a fun and functional platform for unicorn evaluation
What we learned
¡Fuah! We learned how to measure unicornibility of a company using voice and web.
What's next for Unicorn Battle
Train a ML model that would analyze a pitch, landing page and social startup networks like crunchbase to give a score of unicornibility for a company.
Built With
- d3.js
- javascript
- local-storage
- natural-language-processing
- voice-recognition
- webkit



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