SpendSence
Well you can't ask me money when you spent it. So lets make it a Habit to save when you can. We track and Help you come on Track.
Created on 5th July 2024
•
SpendSence
Well you can't ask me money when you spent it. So lets make it a Habit to save when you can. We track and Help you come on Track.
The problem SpendSence solves
This Project Aims to Solve a very serious Problem the people are facing is that they dont have money when they need it, and if we ask them where did it go they don not know because its at your fingertips. In the todays's world everyone likes things fast asnd the same goes with Money but the problem here we are focusing on is about UPI transaction which would help us track your transaction based on your budget.
Anyone who has tried handling his or her own financial situation in today’s fast-paced economy would agree to this fact. Currently, one of the biggest challenges that people struggle with is having no money at the time when they need it, or sometimes people have money, but they do not know where it is going. This concern comes up because of the ease and time it takes to use the UPItransactions where one does not even have to pinch or think before parting with his money. Consequently, it becomes difficult to monitor spending, end up incurring various expenses that are difficult to manage, thus resulting in financial strain.
It signify that the goal of the proposed project is to address several concerns.
This is a very important problem and this project is devised to address it by developing a solution to enable the users to monitor their UPI transactions in accordance with their budgets.
Continuously logging all the UPI transactions performed by the user right from initiating until complete payment.
Grouping these transactions into various expenditure types such as Eating out, Petrol/Gas, Shopping/Accompaniments, Groceries/Clothing, Entertainment/Night out, and Other Large Expenditure.
Creating elaborate expenditure statements and trends in expenditure patterns over time.
Using the motives of expenditure to provide intelligent suggestions on how money ought to be spent in the most efficient way.
Making certain that users are able to have a clear view of their position at any single moment in time, therefore, preventing unpleasant surprises.
Challenges we ran into
Creating SpendSense was not an easy task but was faced with some difficulties. The project was accompanied by a number of challenges ranging from data problems to model building difficulties.
Data Extraction and Cleaning:
Among the problems that arose at the very beginning of the work, the most critical was the extraction and parsing of the information on UPI transactions taken from SMS. The non-transactional formats of SMS coupled with the differing structure of field and the fact that it was not all clean data needed better parsing and data cleaning protocols.
Categorization Accuracy:
It is always difficult to categorize transactions at the same level based on the textual descriptions. This was due to the fact that there were a lot of descriptions that were ambiguous and a lot of merchant names which were not standardized making it very hard to get high accuracy in categorization. Training and evaluation that aimed to activate machine learning models that are capable of studying and predicting categories was also equally bulky.
Real-time Processing:
The main technical concern was to make sure that the transactions in SpendSense would be processed real-time, and the feedbacks be provided to the users real-time as well. This comprised of work done on the above data pipeline to ensure that the backend can also fully support a large number of transactions.
User Interface Design:
Tracks Applied (1)