Madhurima Saha has shared her corporate experience as a Risk Analyst at HDFC and provided suggestions for freshers looking for a role.
Hi everyone. Hope you’re all doing well.
I am a 2019 batch postgraduate from the Department of Statistics, University of Delhi, currently working as a Risk Analyst at HDFC Bank. I am currently posted in Mumbai, working in the Data Science team of the Risk Analytics Unit of the bank.
At HDFC Bank, my work involves building a robust lending mechanism to minimize the losses incurred. I work on the Retail Assets side of the bank’s lending portfolio. Primarily, my work is to build models which can be used to assess the creditworthiness of customers, and estimate income, among various other things. Instead of relying only on the credit scores given by the credit bureaus, like TU CIBIL, Equifax, Experian, and HighMark CRIF, the bank chooses to do an internal scoring of its customers by leveraging the large customer base of the bank. These scorecards are built across all retail asset products, and depending on the data that is being used is of multiple types, like behavioral, application, fraud, to name a few.
The most commonly used technique for these scorecards is Logistic Regression, however, in recent times, Machine Learning models are also being used to build these scores. The tools that I use on a day-to-day basis are SQL and Python.
Apart from building these models, the work also involves conveying to business stakeholders the models’ results and impact. Being able to communicate these with clarity and precision is of utmost importance. Since most of the Policy stakeholders belong from non-technical backgrounds, it becomes important that we can convey to them our findings in the simplest way.
Listing down some of the points which I feel might be of help to Fresher’s who are looking to build a career in Analytics:
- Building a model is a journey of which only the last 5% involves using Machine Learning or other modeling techniques. The majority of it requires working on ETL (Extract, Transform, Load), EDA (Exploratory Data Analysis), and Feature Engineering. Therefore, instead of jumping into building a model directly, I would suggest you take time and familiarise yourself with the data. Concentrate on learning the various ways to manipulate data, and the ways to generate features. It will improve help your model performance.
- While trying new modeling techniques, do not forget that sometimes with the right features, and parameters, a Logistic/Linear Regression model can outperform Machine Learning models.
- To know how to use MS-Excel is extremely important. MS-Excel provides various functionalities which can give you insights into your data. For instance, you can use pivot tables or charts.
- Work towards making your basics strong. Spend time on building basics of probability and statistics, statistical distributions, hypothesis testing, etc. These are important because most of Machine Learning is built on these core concepts. Knowing to code it, but not knowing the concepts may help you in the short run, but will definitely pose problems in the long run.
- Try to secure an internship. Above all, it helps you get a detailed understanding of the work.
- Use online platforms like Coursera, Udemy, etc. to build on SQL, R, and Python skills. There are various other websites like AnalyticsVidhya, Medium, TowardsDataScience that have innumerable informative blog posts on various topics, and the developments in the field. You can also use websites like Kaggle and Hackerrank to practice coding.
Last Piece of Advice
Do not worry if you are unable to master these skills. Every organization has training programs that will teach you, and equip you with the required skill sets. You will get innumerable opportunities to learn new things along the way. Do not be afraid or embarrassed to ask for help.
Lastly, I would just like to say that do not lose hope if things don’t go your way. Just keep learning, keep growing, and you will be there one day. Stay safe, and do not stop believing in yourself.
For any suggestion, please reach out to us on LinkedIn.
Find some of the resources that helped us here.
You can create an impact by talking about your interview experience. Please fill this form .
You can read other articles here.
Cheers and Best!