Data Analyst Research Assistant

Data Analyst

SUMMARY

Enthusiastic data analytics professional with 6+ years experience, possessing a prolific track record of driving customer investment growth, boosting customer retention & enhancing branch business by 40% over 3 financial years (2015-2018) by identifying the target customer. Adept at performing deep dive to gain actionable insights to benefit key high net worth customer & facilitate sound decision-making while generating an error-free report. Proficient in data storytelling to deliver compelling business value to clients & successfully execute projects.

KEY SKILLS

• Data Analysis • Data Exploration

• Data Storytelling • Project Management

• Data Visualization • Client Management

TECHNICAL SKILLS

Languages : Python, SQL, T-SQL

IDE : Jupyter Notebook, Pycharm

BI Tool : Tableau, QlikView, SAP business object, Power BI Cloud : Amazon AWS (EC2)

ETL Tool : SSIS, SSRS, SSMS

SCM Tool : Git

Database : Oracle 10g/11g, MySQL, MS Excel (Advance) PROFESSIONAL EXPERIENCE

Intern - Machine Learning Nov '22 - Dec '22

iNeuron Intelligence Pvt Ltd

Project: Customer Segmentation Domain: E-commerce retail Technology: Python, Tableau IDE : Jupyter Notebook Tools: Python, Exploratory Data Analysis,Cluster Analysis, Flask Framework, Heroku app, Git, Tableau Research Assistant in Machine Learning May '20 - Jun '22 Liverpool John Moores University

Domain: Finance Technology: Python, Tableau IDE : Jupyter Notebook, PyCharm Project : AN EXPLAINABLE COST-SENSITIVE CREDIT RISK ASSESSMENT MODEL Intern - Machine Learning Sep '21 - Oct '21

California

California

Bhawna Gupta

+1-341-***-**** aduoen@r.postjobfree.com Portfolio Fremont, CA Bhawna Linkedin Built a machine learning pipeline for customer segmentation based on their purchasing behaviors using RFM analysis. Examined the performance of K-means, hierarchical and DBScan and K-means achieved better performance results with a silhouette score of 0.50.

Developed a dashboard using Tableau for performance analysis. Deployed the model using the Flask framework

Researched and collected data through literature reviews of 60+ research while accurately cataloging citation information. Led a team of four to schedule and coordinate research activities. Constructed a three-tier machine learning pipeline, and exploratory data analysis for data preparation, model building, and training.

Designed a Stacking model of boosting classifier (XGBoost, LightGBM, and Catboost) to handle imbalanced data and reduce the misclassification cost.

Developed model with SMOTE technique XGBoost and Stacking model achieved the highest performance with about 99.86% AUC Score, 99.96% F1, and 99.86% G-mean.

Performed A/B testing to reject the null hypothesis that boosting the ensemble model with oversampling technique reduces the misclassification cost.

Integration of explainable AI SHAP achieved the interpretation of the prediction result that can be explainable to borrowers. iNeuron Intelligence Pvt Ltd

Project : Restaurant Rating Prediction Domain: Retail Technology: Python IDE : Jupyter Notebook, Pycharm Tools: Python, Exploratory Data Analysis,Predictive Analysis, Flask Framework, Heroku app, Git, SSIS,, Tableau Intern - Data Analytics Jun '21 - Jul '21

Technocolabs

Project : Loan Risk Prediction Domain: Banking Technology: Python IDE : Jupyter Notebook Tools: Python, Exploratory Data Analysis,Predictive Analysis, Flask Framework, Heroku app, Git, SSMS Financial Data Analyst Officer Jun '15 - Sep '19

Canara Bank

Domain : Banking & Finance Technology : SQL, oracle 10g, Microsoft Excel, SAP Business Object Tools : SQL, Oracle 10 g, Microsoft excel,SAP business object. MS Project, SSIS QlikView Developer (Data Analyst) Jun '13 - Jun '15 QUOSPHERE

Domain: IT Technology : SQL, Qlikview

Tools : QlikView, Tableau, SQL, Oracle 10g

EDUCATION

Master of Science in Data Science May '20 - Jul '22 Liverpool John Moores University

California

California

Bengaluru,Karnataka, IN

Mumbai, Maharastra, IN

California

Performed ETL(Extract Transform Load) using the tool SSIS to integrate with jupyter notebook to run the SQL script to transform the dataset for the machine learning model.

Performed extensive Exploratory Data Analysis to clean up the dataset afterward univariate and bivariate analysis was done. Constructed an ensemble model of Decision Tree, Random Forest and XGBoost and the result was compared with a multiple regression model.

Achieved the highest R squared score of 93.87% for the train set and 80.98% for the test set using the ensemble model after performing A/B testing.

Developed a sales dashboard for rating prediction using Tableau. Extracted data using SQL Server Management Studio(SSMS) and performed data manipulation, and aggregation using DDL, and DML SQL queries and then loaded data to jupyter notebook to perform data wrangling before hypothesis testing. Designed a prediction model using Random Forest after preparing the data using exploratory data analysis. Oversampled the imbalanced dataset using SMOTE technique and achieved a 71.2% FI score for the test set. Using Heroku app launch a website (designed with HTML/CSS) to take the input from the customer and show the loan approval decision.

By analyzing and interpreting data and making comparative analyses, proposed changes in methods to identify the target customers and improve the branch business by more than 40% over 3 financial years. Evaluated the creditworthiness of the customer to determine their financial health to repay the credit which leads to reducing the NPA by more than 5% of the branch.

Identify and drive process improvements, including the creation of standard and ad-hoc reports, tools, and Excel dashboards. Involved in ETL development, creating required mappings for the data flow using SSIS. Performed extraction and transformation of data from multiple data sources and created a financial dashboard for better insight using Excel.

Computed multiple KPIs balance sheets, cash management, and Revenue generated from credit to track the branch business. Design the personalized investment portfolio dashboard for the HNI (High Net worth Income) customer based on their future goal.

Improved the number of clients for Quosphere by delivering onsite 'Proof of Concept' QlikView dashboard with a detailed demonstration by reviewing project requirements to identify customer expectations and resources needed to meet goals. Generate standard or custom reports by identifying the KPIs of business, financial, or economic data for review by executives, managers, and clients.

implemented the 3-tier ETL architecture for data extraction, modeling, and application and stored it in the QVDs. Created an automated mailing system to generate the daily sales reports status branch-wise to know the business target position by implementing incremental loading using QVDs. Percentage : 71%

Bachelor of Technology in Computer Science and Engineering Aug '07 - Jun '11 BBS College of Engineering and Technology Allahabad,UP,India equivalent WES score : 3.95/4.00

Percentage : 75%

Contact this candidate