One of the E learning platform facing problem to target their customer by using conventional LEAD algorithm as
precision of the algorithm is only 20%. They want us to make a predictive model which can able to predict
retention of a customer precisely.
Analyze the attribute of data very deeply to improve the performance of the model.
Handle a very large amount of imbalanced data.
Improve result by 37.5% as compare to the conventional Lead algorithm.

nalyzing the data extensively, on multiple levels and aggregations; drawing inferences from the same and to
make it suitable for predicting the quality of candidature of a loan applicant.
Work on a bunch of data analysis to find scope for improvement
Co-ordinate with global teams to understand their requirements, understanding the problem working alongside
them
Collaborate with the team in writing and testing highly efficient, optimized and robust code that is easily scalable
Collaborate with Data Engineers, Data Analysis and the Data Science team on the design, development and
maintenance of the product
Communicate findings and obstacles to team to achieve best approach

One of the large scale Medical insurance company was facing a problem of losing costumer because of taking
too much time to repayment claim amount.our client want us to make a predictive machine learning model
which predict repayment amount based on documents(Discharge summery, Invoice) provided by hospital.So
they can repayment claims to costumer in time.
Handle Discharge summery document by using NLP techniques.
Apply Text processing and document classification techniques to extract useful information from given
document.
Make a data frame from the document so that it is useful for our model.
Worked with the Deep Learning team to build predictive models