This image is currently processing. Please be patient...
While still going to school, we have been inspired by many giants to work on our own startup. The cycle of generating, pivoting, and moving on with ideas happens on a weekly basis.
Coming up with new startup idea, we often could not foresee what lies ahead. We lacks the experience of a successful entrepreneur. Thus, Edia was created. This application aim to lend a hand and allow entrepreneur and maker to have a deeper understanding of their next big thing.
We believe that once an idea was chosen, the user will need to go out and research all over the places. They will read and absorb new information. And years from then, they will become a new person, with personality crafted up from all the reading, experimentation and knowledge gaining from doing/researching their startup idea.
So, we believe by analyzing your possible future self (who will be throwing their time and money into that startup idea) you will be able to choose whether you like this outcome or not. Using Watson to analyze a broad range of document related to their ideas, and analyze the writing of those idea's abstract, we get a personality that might represent the founder in the future.
We call this the emotion test for your idea.
What it does
From user's 5 seconds pitch often in the form of "a product for a niche market," we extracts all the keywords using Watson's Alchemy language API.
For each keyword obtained, we then run them through Watson's Concept Insight to get 36 more related concepts (9 each level, from 0 to 3) and their abstracts.
We combine all of the concept abstract into a giant body of text (which are often 5000-6000 words), and submit it to Watsons' Personality insight, and retrieve back the "emotion" of this startup idea.
We also run the raw 5 second pitch through Watson's Tone Analyzer and return a summary of their tone.
How we built it
- Front end: Ionic, Angular and Sass
- Back end: Nodejs, Restify hooking up with all of IBM Bluemix services
Challenges we ran into
- We had literally 24 hours to work on this hackathon. We started late on the 11th.
- Representation of the data was very challenging as we had to study the data format
- We got into a point where we need to make multiple async call to Watson. That was the first of us to handle it in nodejs.
- Also our first time with data chart and visualization on the front-end!
Accomplishments that we are proud of
- Integrating all of the Watson's services and have them run flawlessly
What's next for Edia
- We want to take in the user's social media content (twitter/facebook), run it through the personality service and compare it with the idea's personality to see if they are the right person to found it.
- We also want to find them a co-founder, or someone they should consult for their idea.