Eisha Mahajan

682-***-**** ac97wz@r.postjobfree.com

linkedin.com/in/EishaMahajan

EDUCATION

The University of Texas at Dallas, USA May 2019

Master of Science in Information Technology and Management GPA 3.3 TECHNICAL SKILLS

Programming: SQL, Python (numpy, pandas, matplotlib, seaborn, scikit-learn, keras), Spark, Hadoop, Hive Machine Learning: Linear/Logistic Regression,SVM, Decision Tree, Ensemble models, PCA, Clustering, Statistical Tests Analytical Tools: MS SQL Server, MS Access, Tableau, Power BI, Google Analytics, SSIS, AWS EXPERIENCE

Audit Data Analyst Intern Bank of America June 2018 - Aug 2018

Upgraded from MS Excel to Tableau which helped investigate large sets of data, facilitated trend & anomaly detection, helped bucket high & low risks which improved risk assessment by 10%

Broke down business problems into key achievable goals which helped uncover 20% issues early in stage

Created SQL scripts to perform data validation which saved manual effort and time by 20% Systems Engineer Tata Consultancy Services Sept 2014 - July 2017

Analyzed historical defects in JIRA to identify their frequency and pattern, which helped optimize testing strategies and reduced cost of escaped production defects by 25%

Transitioned from exhaustive regression testing to identifying critical user requirements which provided 80% test coverage thereby saving 40% manual effort and time

Performed data reconciliation through SQL stored procedures, triggers which improved the test efficiency by 20% PROJECTS

Fraudulent Firm Classification - External Audit

Led the development of a model capable of classifying fraudulent firms with 98% accuracy and 0.97 AUC in scikit- learn for an external audit company

Conducted EDA in python to uncover outliers using IQR statistic and dropped columns having multicollinearity

Evaluated several models using GridSearchCV in which kernelized SVM outperformed. Optimized to achieve a 20% rise in Recall score which would aid in dynamic audit planning Building Dream11 Soccer team

Performed EDA on 17k records to discover distinctive insights about player’s position, market value, potential, top features and club in pandas & seaborn

Developed a regularized model using Linear Regression with 93% R2 to predict player rating that helps identify important features and aid in building best teams and designing winning strategies Instacart Market Basket Analysis

Investigated data having 3M grocery orders to discover customer’s purchasing patterns, frequency of orders, reordered products, and high demand products in python

Implemented Apriori to build frequent item pairs and applied association rules with 0.01% support to find the top items likely to be purchased together which can serve as recommender system to improve and optimize sales How Uber leveraged data to expand

Created a narrative story in Tableau using interactive dashboards about the rise of Uber in Manhattan

Visualized correlation between rides, neighborhoods and precipitation; categorized rides by months, weeks and hours to observe how Uber leverages data to predict demand and supply, set dynamic fares and optimize services Is Chicago’s 311 potholes complaint department efficient?

Examined 311 complaints on potholes to uncover how government works behind the scenes using pandas, seaborn

Engineered datetime fields to create/extract features which enabled complaint creation & resolution time comparison by year, month, districts, streets

Discovered inverse relation between complaints & resolution which uncovered request handling strategy. Districts with worst conditions had exponential drop in complaints from 2010 - 2017

Contact this candidate