Summary
Overview
Work History
Education
Skills
Languages
Additional Information
Timeline
Generic

Maheesh Dubey

Gurugram

Summary

Results-driven Data Scientist with 5+ years of experience in fraud analytics, risk management, and advanced data science applications within the financial services sector. Proven track record in developing predictive models, implementing fraud detection strategies, and leveraging cutting-edge technologies to drive operational efficiency and mitigate risk. Skilled at analyzing complex datasets, identifying patterns, and providing actionable insights to optimize business processes. Committed to staying ahead of industry trends and continuously enhancing skills to deliver innovative solutions for evolving data science challenges.

Overview

5
5
years of professional experience

Work History

Lead Assistant Manager

EXL Service.com (I) Pvt. Ltd
11.2023 - Current
  • Fraud prevention Data Model creation:
  • Architected comprehensive SQL-based data extraction processes for transaction history compilation
  • Developed robust query mechanisms to support advanced fraud prevention modeling
  • Facilitated secure data transfers to external risk analysis partners using Git version control
  • Key Achievement: Enhanced fraud detection model resulting in a 22% improvement in detection accuracy
  • Automation of Fraud operations:
  • Designed and implemented Python-driven automation for account update processes
  • Integrated multiple fraud detection platforms including IOVATION and EMAILAGE
  • Significantly reduced manual intervention and operational redundancies
  • Impact: Compressed process timelines from 3 days to 2, generating annual time savings of 48 working days
  • Application-Level Risk Assessment Expertise:
  • Implemented multi-dimensional risk scoring methodologies using third-party vendor assessments
  • Developed sophisticated risk evaluation frameworks for comprehensive customer base analysis
  • Result: Increased fraud detection rate by approximately 12% through advanced scoring techniques
  • Specialized Fraud Detection Systems
  • Engineered ACH Mismatch Detection System to identify complex deposit irregularities
  • Developed analytical models for detecting suspicious non-monetary transaction patterns
  • Outcomes:-
  • Identified and closed approximately 3,000 potential money laundering accounts
  • Detected and mitigated 2,500 fraudulent accounts through advanced pattern recognition
  • Strategic Monitoring and Visualization
  • Created comprehensive monitoring dashboards using Tableau and Amplitude
  • Implemented real-time fraud pattern tracking systems
  • Enabled agile response strategies through advanced data visualization techniques.

Data Analyst

Molars Technologies Pvt. Ltd
10.2019 - 11.2023


  • Portfolio Analysis, Reporting and Strategy Building on default loan prediction
  • Data Modeling and Data Integration:
  • Use SAS, Python and SQL to construct data models that consolidate and integrate this data into a DataMart
  • This DataMart should serve as a central repository for loan origination data
  • Automated Reporting and Dashboarding:
  • Apply your programming and reporting skills to automate the generation of reports and dashboards
  • Use data science tools like Excel, SQL, Tableau, Alteryx, Python and SAS to set up automated processes for data retrieval and visualization
  • Data Modeling and Loan Default Prediction:
  • Predict whether a customer is likely to default on a personal loan
  • Objective: Lower loan default rates, reduce financial losses, and improve decision-making
  • Collect data on past loan applicants, including both those who defaulted and those who did not
  • Data sources may include internal records, credit bureau data, and other relevant sources
  • Collect data on past loan applicants, including both those who defaulted and those who did not
  • Data sources may include internal records, credit bureau data, and other relevant sources
  • Handle missing values, duplicate records, and outliers
  • Create relevant features such as debt-to-income ratio, credit utilization, and payment history
  • Explore the data to gain insights into the loan default patterns
  • Visualize relationships between features and the target variable (default or non-default)
  • Identify correlations and trends that may influence loan defaults
  • Split the dataset into training and testing sets to evaluate model performance
  • Choose appropriate machine learning algorithms for classification tasks
  • Common choices include Logistic Regression, Random Forest, Gradient Boosting, and Neural Networks
  • Consider experimenting with multiple algorithms to find the best-performing one
  • Train the selected models using the training dataset
  • Tune hyper parameters using techniques like grid search or random search to optimize model performance
  • Evaluate the models on the testing dataset using appropriate metrics such as accuracy, precision, recall, F1-score, and ROC AUC
  • Consider using a confusion matrix to understand false positives and false negatives.

Education

BTECH -

Medicaps Institute of Technology And Management
Indore, India
2019

Skills

  • Fraud Detection & Prevention
  • Predictive Modeling
  • Machine Learning Algorithms
  • Risk Assessment
  • Process Automation
  • Statistical Analysis
  • Data Visualization
  • Database Management
  • Programming Languages: Python, SQL, SAS
  • Data Analysis Tools: AWS, Excel, VBA, Power BI, Tableau, Amplitude
  • Version Control: Git
  • Fraud Detection Platforms: Iovation, Emailage
  • Specialized Tools: Visa RSM, LexisNexis, MS Access

Languages

  • Language Known -English, Hindi
  • Additional Information

    • DOB: 27/10/1997
    • Gender: Male
    • Nationality – Indian
    • Permanent Address - 41/17, Jangipura, Dabra, district-Gwalior, Madhya Pradesh, Pin-475110

    Timeline

    Lead Assistant Manager

    EXL Service.com (I) Pvt. Ltd
    11.2023 - Current

    Data Analyst

    Molars Technologies Pvt. Ltd
    10.2019 - 11.2023

    BTECH -

    Medicaps Institute of Technology And Management
    Maheesh Dubey