Smart Stocks is Stock traders dashboard Application created using Django Framework with postgres as the database.
-
Stock prices of over 8000 companies as listed on NASDAQ.
-
Implemented independent jenkins job which update stock prices of all the companies daily at 5:00 PM
-
Implemented independent jenkins job which updates Machine Learning models only of the companies which have been visited by the user. Updates every Monday at 7:00 AM
-
The price predictions generated are real-time. The model takes into account the latest 60 days stock prices to make a dynamic prediction.
-
Recent News of a particular stock are readily available right below the price prediction so that user can make an informed decision based on the current news and trends.
-
Implemented Continous Delivery using Rundeck
-
Implemented Neural Networks with LSTM nodes for predictions using keras. With the structure : 60 - 50 - 50 - 1. The hidden layers are the LSTM nodes. Input and Ouput layers are Dense Nodes.
