Forecast and analyze market trends to minimize losses and in turn maximize gain through investment in cryptocurrencies.

What it does

Provides a detailed insight into the past and the present patterns, predict future fluctuations and notify users when to trade cryptocurrencies based on user budget and time period of investment.

How we built it

  1. Python using streamlit
  2. Historic data of the cryptocurrencies from yahoo finance
  3. Prediction using an open source tool: fbprophet
  4. Plots using seaborn library

Challenges we ran into

  1. Hosting the dynamic application
  2. Google cloud computing machine was too slow for our application. We had to generate the plots from our localhost and add it on the static webpage.

Accomplishments that we're proud of

  1. We are very glad to make Cryptopiens public.
  2. The web application is available for use at The github repository contains the code for dynamic usage of the application.
  3. This dynamic version lets the user select their budget, risk allowance, number of quarters, etc. The user gets an overview of the historic and the current market performances of all the cryptocurrencies supported by us alongwith an app based suggestion depending on their investment time and budget preferences.

What we learned

  1. We used a new machine learning time series prediction model in this project, FbProphet. The library is open source.
  2. We used a github page and linked it with an external domain using a new domain we bought at

What's next for Cryptopiens

  1. Cryptopiens is soon going to support DeSo. We are waiting for availability of more data.
  2. The website only has a static page as the host Github pages only support static pages. We are planning to use google cloud hosting and computing to work around this.
  3. Cryptopiens is soon going to have an implementation of a graph-based risk measurement tool for crypto sapiens.

Built With

Share this project: