Let’s think about pastime we are fond of. Travelling, fleeping news feed, watching movies… Have you got аny ideas?
But how to mix business with pleasure? The answer is food. We cook it ourselves, buy semi-finished products and visit restaurants. Certainly, restaurants existed many many many years ago and are not going to disappear.
That's why our idea relates to this service.
Imagine: you have just appeared at the restaurant door and the staff has already known, what kind of food would please you most of all, while you just had no idea what to order. What if you also get a discount for your food preferences?
We’ve developed a system, combining the features of:
- mobile apps for booking and searching restaurants;
- CRMs.
DealWithMeal provides the further options:
- Automated process of a restaurant visiting (Searching by criteria, booking a table);
- Intelligent recommendatory system, which helps to choose the most desired dishes;
- Discount system (e.g., for favorite visitor's dishes).
You could see our deep research here.
It was very complex process. We had a lot of work to do so it was very important to organize ourselves perfectly. We know each other pretty well, so we interact with each other successfully. It helped us to do everything on time.
What was actually done?
Firstly, we have spent a lot of time brainstorming and developing the main algorithm and architecture of our program.
Secondly, we pointed the aims, were to be reached. It was enormous work to make the main architecture, think about the questions in details of the realization for some features. While we were implementing the project, we faced a lot of problems and bugs, some of them were so stupid, that we have laughed at when they were fixed.
Our system has the main following components:
- Back-end (Django Rest Framework, sqlite3),
- front end:
- website (JQuery, HTML, CSS),
- iOS application (SWIFT).
Our algorithm of recommendatory system uses the mechanisms and techniques of the machine learning. It’s intended to advise dishes for the client in 3 different cases:
- This is client`s first visit to any restaurant connected with our system;
- The client has already ordered something in any restaurant connected to our system;
- The client has already ordered something in the current order.
There is no matter if client order the table at website, by app or just in the restaurant – we will advise him something he would be glad to eat!
Our server was built using REST-architecture, so we have got a good opportunity to deploy our app on all platforms.
For the first challenge, it was to generate an idea. It was very difficult to come up with something new, but finally we've done it. And now we believe in DealWithMeal.
The other challenge was to focus on the work. Junction 2016 has a very multiform program: amazing events, interesting partners and guests, new people. But call of duty made us sit and code.
We also managed to keep good communication channel between the members of our team. It is difficult to reach understanding, when everyone has an unique task.
Our crew is proud of our project. It seemed impossible to perform such a result for 36 hours, but we dealed with it.
We also dealed with developing entire, intelligent system, which can be really useful both for restaurants and visitors.
We are also proud of each other. Every member of the crew is vital, every member is inseparable part of "BORSCH" crew.
First of all, we have learned how to deal with the problems if the time is short. Secondly, we have learned how to separate the duties effectively. We also have drilled the co-working skills and met many interesting people.
Our aim is to implement our project in real life. We are sure that it is very useful.
We would like to extend our recommender system and work not only with dishes, but with restaurants' logistics, purchasing; to improve the algorithm, to make it faster and more intelligent.
- Sergey Ignatushchenko - Back-end developer;
- Alexander Shoshiashvili - iOS developer;
- Maksim Shevchenko - Front-end developer, Analyst;
- Alexander Ivanov - Analyst, Researcher;
- Denis Merkushin - Analyst, Researcher.