Have you ever wanted to trade all your savings? Of course not... But have you ever wanted to trade all your savings without being affected? Maybe... If you are the kind of person who likes to risk and try out a new trading algorithm ideas, Pedlar is the way to go.
The name Pedlar symbolises the way we imagine users trade on our platform – as a one-day trader, they enter, then do their trade and leave.
Motivation & What it does
On a more serious note, our motivation was mainly based on the fact that we hoped to create something we had not idea about. What would be a more suitable field than finance? Building a new way of recruiting algo-trading-programmers for trading companies seemed an absolutely remarkable idea! This new way of making challenges came to us as we have already passed through quite a few interviews and only a couple of them really have stood out. With this idea in mind, we developed a game consisting of programming your own bot for trading and playing in the real market.
How it works
The project is split into three parts: • the user part – the actual bot • the server part – the actual market • the front-end part – the actual interface
We implemented a web API which the users can connect to in order to play the game. There is a base code provided to the users, with which they can interact with our API, that is hosted on the Google Compute Engine. We also built a website, where players can see their balance in REAL time. This again is connected to the server.
It is up to the user to design and implement their own algorithms, potentially by using the code we provide that interacts with our services. We hope that users will get creative, as there are basically no limits to what they can use in their bots, i.e. sentiment analysis.
Challenges we ran into
The hardest part was to connect the aforementioned components together by using VueJS and Flask. Other technologies that we used include MongoDB, Google Cloud Platform, Domain.com. We had in the first place problems with connecting the website with the back-end, making the countdown working on the website, changing the ranking in real time. We did not know where should we take the data from, database-ul was not updating itself.
What is next
We have a lot of ideas to implement, which would make the simulation more realistic: • simulate multiple stock exchanges at the same time • minimise latency • make the market adapt to the users’ actions • make the rounds configurable by potential clients
Accomplishments that we're proud of
All, the challenges we overcome, and the fact that we actually had time for 4-5 hours of sleep.
What we learned
We actually learned how to use Vue.js, MongoDB and some others, neither of us had worked with it before.