Inspiration
Regarding data, most novices will be afraid of stock K-line charts and dense real-time prices. We combine visual data with chatbot. Users can ask InvestBot questions, and InvestBot will analyse the user’s message and finally reply to it. User-friendly Visualisation chart. It is clear at a glance and can be inquired at any time.
What it does
Our product is called InvestBot. The appearance of this product is to help users clearly and intuitively understand his current investment situation and predict and evaluate the next stock price and potential risks for users.
How we built it
On the back end, we used Python to count the data and trained a neural network to predict the next stock price. In the previous paragraph, we trained InvestBot through the English corpus. Our InvestBot is a web version, which is based on Flask.
Challenges we ran into
In the process of doing the project, we encountered some challenges. How to find useful data integration is our first obstacle, we tried our best to search the stock data set in Kaggle through clear data requirements to ensure the matching of data information. The second challenge is that our team lacks a wealth of front-end experience. Understanding Flask and HTML code has become the main bottleneck of our project. With our joint efforts, we finally understood it and found a solution.
Accomplishments that we're proud of
Our neural network was successfully trained, and the test set satisfies our needs very well. We are essentially back-end engineers and have no front-end development experience. We successfully wrote the front-end and connected with the back-end, which is very exciting.
What we learned
Our chatbot is based on Flask to implement the web version. These are all new knowledge for us. We learned in the process of work and gained experience in front-end development. After seeing our project finally functioning normally, tears fell.
What's next for Investment Robot for blackRock
We will call more functions. In fact, we have more prediction functions. We only chose the best one trained with a neural network. Similarly, the training set of our chatbot can also be further optimised to obtain a better user experience. If time is abundant, we can guarantee to make products with excellent interaction.
Log in or sign up for Devpost to join the conversation.