BärSaver - 365º visibility on expenses, effortlessly finances management
The inspiration to build the BärSaver came to us when we were working in Vietnam 2 years ago. Foreign country, currency was one of the most important factors which affected on our spending. There were difficulties on understanding on what and how we did spend our money.
People around had similar issues, and the income level was not the main issue. However, one of most important points for our team was and is to guide low-income citizens to manage their finances better and effortlessly.
You’ll be surprised how life-changing something so simple can be. When you’re on top of your money, life is good.
We help you effortlessly manage your finances in one place.
What it does
BärSaver - is the product to improve customers financial situation, by setting goals, giving 365º visibility on expenses.
The application will give insights and data to boost spending behaviour of a user.
BärSaver automatically categorizes your transactions, application uses machine learning algorithms to constantly improve classification.
We also use Deutsche Bank API, to enable us to access directly into customers bank account to process transactions. We have tens of default categories, and it's easy to re-categorize any transaction as you please.
How we built it
Idea -> Team -> Development -> Failure -> Food -> Development -> Great Success
Challenges we ran into
There were a number of challenges we run into, both product and technical. The main challenge for us was a time given. We needed to come up with feature set which is possible to implement in 21 hours (from 12:00pm till 09:30 am). We tried out a lot of applications which claim to improve money management. Most of tested services are not doing enough, and most importantly require manual job to be done.
We wanted our application to do most of the duties automatically so user will not be required to do data entry for all the expenses.
Accomplishments that we are proud of
BärSaver, classification algorithm is very efficient. We accomplished better auto categorization than Deutsche Bank’s financial planner. In our tests 70% of transactions were not categorized by Deutsche Bank, which does not give any visibility on spending’s and planning the future budgets.
Our application has goals and recommendations on optimization of costs and budgets.
What we learned
- Time given is never enough!
- Nothing goes as it has been scheduled
- Deutsche Bank API products: AccountInsights and CustomerInsights
- Symfony 4, a high-performance PHP framework and a set of components
- scikit-learn: machine learning in Python
- Maksim Naumov (linkedin, github)
- Vladimir Bolshakov (linkedin, github)
- Sardorbek Pulatov (linkedin, github)
- Sonya Bolshakova (linkedin)
What's next for BärSaver
- Bill recognition and categorisation of cash withdraw transactions
- Better classification algorithm
- Better recommendations on optimizing spending