Lovely Trolly

Inspiration

There are trolls on a lot of platforms. But what if we automated the way we entertain them.

What it does

Lovely Trolly is an API that takes in your desired subject and the text of the troll that you would like to respond to... It returns text that sounds like it would be in the language you are writing, but not really.

How we built it

Most of the design is a simple plan on Sketch. Given the text inputs mentioned above. We scrape the web for articles and feed it to the "language learning model" written in Go.
simply: Design API endpoints. Create the data. Write a handler to add a new text query. Write a handler to respond with a paragraph or 3.

Challenges we ran into

So far, I haven't gotten far. I only have the beginning of the Markov-inspired program that learns based on the articles of the subject of interest and the troll's text.

Accomplishments that we're proud of

I'm proud of coming up with such a funny problem and solution.

What we learned

Don't start too late to begin testing.

What's next for Lovely Trolly

Extend the ability of the core learning mechanism.

Built With

• go