BulleTrades

Features

Authentication Using Bcrypt Hash to personalize each of our users
Users can manage each trade with real-time data to track and prevent false transactions
Users can purchase and sell tickers with partial shares and dollar amount (backend algorithm)
Users can create, read, update, and delete watchlists with an ergonomic UI design
Modern searchbar and query feature that color coordinates two sets of results to allow users to find their desired result (prevents "tip of the tongue", lethologica)

Background

BulleTrades is a full stack web application inspired by Robinhood. BulleTrades allows users to personalize their account by having them create an account and protect them with a session token and password that is never saved or displayed to other users. As a customer, you have the ability to search for any ticker and make real time buy and sell orders that will impact your overall buying power. Users also have the ability to view real-time information about a ticker before placing a trade giving them the power to analyze a stocks worth. Visually, the site incorporates a live chart that users can hover over and check their progress for each stock and their respective portfolio. --This project was built in 2 weeks

Technologies

Frontend

Javascript
React / Redux

Server / Router / Controller

Jquery ajax
Ruby on Rails Router

Backend

Ruby on Rails Associations
PostgreSQL

Website Features

Dynamic SearchBar

Bulletrades incorporates a dynamic search bar that gives users a color-coordinated search result to pinpoint their desired result. All results are non duplicate to prevent additional queries

Purchasing and Selling both Partial Shares and Dollar Amounts

Bulletrades not only keeps track of each user purchases to prevent invalidated transactions, but we also implemented a binary search algorithm that finds the closest value to the user input price. Then it proceeds to delete objects starting from the index with two pointers with an O(n) time complexity.

def self.use_binary_search_to_find_closest_index(array_of_objects, avg_ticker_price)
        if array_of_objects.length() < 3
            return [low = 0, high = 1]
        end
        if (avg_ticker_price < array_of_objects[0].avg_ticker_price)
            return [low = 0, high = 1]
        end
        if (avg_ticker_price > array_of_objects[array_of_objects.length() - 1].avg_ticker_price)
            return [low = array_of_objects.length() - 2, high = array_of_objects.length() - 1]
        end

        low = 0
        high = array_of_objects.length() - 1
        while (low <= high) 
            midpoint = (high + low) / 2
            if (avg_ticker_price < array_of_objects[midpoint].avg_ticker_price)
                high = midpoint - 1
            elsif (avg_ticker_price > array_of_objects[midpoint].avg_ticker_price)
                low = midpoint + 1
            else    
                low = high - 1
                return [low , high]
            end
        end
        return [low, high]
    end

Create, Read, Update, Delete Watchlists with Modern UI and Fast Backend Query

Watchlists and Tickers are a many to many relationship, I used a joins to table to reference and update each other accordingly.

Client Caching to prevent repetitive and unnecessary user requests to API

export const fetchNewsByTickerName = (tickerName) => {
    let newsData = sessionStorage.getItem(`${tickerName}`)
    if (newsData !== null) {
        return new Promise((resolve, reject) => resolve($.parseJSON(newsData)))
    } else {
        return $.ajax({
            url: `https://gnews.io/api/v4/search?q=${tickerName}&max=5&lang=en&country=us&sortby=publishedAt&to&token=${apiKey}`,
            method: "GET",
            success: function(data) {
                sessionStorage.setItem(`${tickerName}`, JSON.stringify(data.articles))
            }
        })
    }
}

hankc97/Bulletrades