Default profile banner
CM

Chayan Mehrotra

@chayanmehrotra

Senior Consultant at Deloitte USI

Gurgaon, Haryana

Deloitte USIScaler Neovarsity in collaboration with Woolf University

Chayan Mehrotra is a Data Science professional with 6 years of overall experience, including 4+ years in Data Science and Analytics. He has extensive experience working on client engagements for major US Banks, focusing on risk modeling and predictive analytics. His technical expertise includes supervised and unsupervised machine learning algorithms, hyperparameter tuning, and developing solutions using Python libraries like Scikit-Learn, Pandas, and NumPy. He is skilled in areas such as Counter Party Credit Risk, Hypothesis Testing, and deploying models using Streamlit.

Experience

Senior Consultant (Data Science)

Deloitte USI

•Feb 2001 - Present

Worked on Logistic Regression Model to predict potential default customers for a loan application. Experience in Model Risk Management Documentation according to SR 11-7 Guidelines. Retail Analytics projects for a French Retailer to migrate Private Label Cards to Co-Branded Credit Card. Created and Deployed an Auto-ML product on Streamlit to detect the classifiers based on Algorithm Selection. Interpreted ridge regression model & LASSO solver via gradient descent to select the regularization parameters. Independently revalidated Counter Party Credit Risk Statistical Model using Hypothesis Testing Proof-of-Concept. Strong understanding of Statistics(Hypothesis Testing (ANOVA, T-Test, Z-Test, Levenes Test, Box-Cox,Shapiro). Montecarlo Simulation to generate random 1000 paths for backtesting. Achieved a Slack Coefficient of 1.0 along with Radial Basis Function using GridSearch CV Algorithm. Trained 3 team members on basic SQL ,pandas and POC ML Algorithms. Worked on Ensemble Techniques such as Random Forest Classifier, XG Boost along with Optimization Algorithms.

Analyst

NSE Indices, Mumbai

•Jun 2001 - Jul 2001

Developed a VBA code to calculate price of a T-Bill.

Operations Sr. Analyst

IHS Markit

•Aug 2001 - Mar 2001

Worked on Data Analysis using SQL and Tableau to extract key business insights. Used SciPy for hypothesis testing to test claims. Developed Yield curve simulator in Python using pandas to generate scenarios for upwards and downward movement of Interest Rates. Worked on a live project for predicting bond returns using SVM Regressions.

Business Analyst

Evalueserve

•Apr 2001 - Feb 2001

Worked on Support Vector Machines algorithm to build a challenger model to compare Logistic Regression. Campaign Analytics project for a Bank (Model Development and Validation) [Marketing Campaign for Annuity Project]. Predicted increase in additional revenue by targeting only 40% of the customers through LOGISTIC REGRESSION MODEL. Developed a Python Class to calculate Market Risk parameter (Value-at-Risk, using various Significance Levels). Worked on Random Forest Classifier to predict the customer subscription rate for a Term Deposit Product. Developed SQL query to extract insights from index prices for various asset classes. Delivered visualizations in Python using Seaborn, Matplotlib libraries.

Education

Scaler Neovarsity in collaboration with Woolf University

Masters in Data Science with Specialization in AI/ML

Jun 2001 - May 2001

MPTSME, Narsee Monjee Institute of Management Studies

Dual Degree MBA ( Technology Management Finance ) B.Tech (Electronics and Telecommunication)

Jun 2001 - May 2001

Extracted and presented key business insights to the management for various Asset Classes. Developed a Valuation model for pricing of plain vanilla bonds in Python. Used Clustering to identify potential Fund Categories.

Skills

Scikit-Learn
NumPy
SciPy
Plot.ly
Pandas
Matplotlib
Statsmodel
seaborn
openpyxl
scikit-optimize
Oracle
MySQL 8.0
Linear Regression
Logistic Regression
SVM
Decision Trees
Random Forests (Bagging)
Boosting
Clustering
Gradient Boosted trees
Hypothesis Testing
Linear Algebra
Excel
Tableau
Model Deployment -Streamlit
K-Nearest Neighbor's
K-Means
DBScan Clustering
Data Analysis
Statistics
Supervised and Unsupervised Machine Learning
Dashboards (Tableau)
Model Evaluation (Bias - Variance Tradeoff)
Cross Validation
XGBoost
VaR
Monte Carlo Simulations
Counter Party Credit Risk
Backtesting Methodology
Bayes Theorem based Classifier
Bayesian Optimization