Feather | Ethical Investing, Made Easy

An AI bot that helps ethically-conscious investors find companies that fit their moral code.

Comment

Inspiration

I became inspired to create Feather when I began trading stocks with some of my spare cash and getting involved in investment. I came to realize how many companies I see as unethical, and wanted to create a way for people like me to find investments they can be proud of.

What it does

Feather uses cutting-edge AI technology to enable socially conscious investors to both quickly research stocks and find stocks that align with their values. At current, the main two features of Feather are to analyze a list of stocks (given by the user), and suggest stocks to the user, both based off of a short, quick questionnaire designed to map the user's moral universe.

How we built it

Feather originally began as a full-stack web application utilizing node.js and the Google Gemini API, but quickly changed into a static website implementing the Deepseek API. It was built from the ground up in HTML5, CSS, and Javascript.

Challenges we ran into

As mentioned before, Feather began it's life as a node.js application, but it became clear within a few hours that the team (myself) lacked the skills to implement a full-stack web application. The choice at this point was to stick to our guns and try desperately to use Gemini (which required node.js,) or bite the bullet and use a different AI model. Ultimately, due to time constraints, the switch was made, which is why Feather is built on Deepseek.

Accomplishments that we're proud of

Firstly, getting a project complete at all is something to celebrate! But aside from that, we're proud of getting an API to work, of making a functional AI app, of making something with the potential for genuine positive impact if developed upon, and most of all, we're proud of how much we accomplished with such a small team (1 person).

What we learned

The team (again, me) learned a lot about API usage in the process of implementing the Deepseek API. I learned many functions of CSS and HTML in the process of making this web app as well. Additionally, despite ultimately abandoning the full-stack build of Feather, we learned a lot about how a frontend communicates with a backend, and vice-versa.

What's next for Feather

A feature we toyed around with including in the RevolutionUC build of Feather that we would love to include in the future is implementation with the Robinhood API. We're not familiar with how extensive it is, so this might be a fool's errand, but we want a feature for investors to be able to import their own investment portfolios into Feather. Other key features we plan on including are the ability to save a user profile and values (either via cookies or a proper backend) to expedite the process, and mobile support with a reactive design.

Built With

Share this project:

Updates