Still Struggling with Tech Hiring? Discover Olibr's Solution Now!
Login / RegisterDeveloperClient
Login/Register

Rajeshwari M

SENIOR DATA SCIENTIST

  • Bangalore Urban, KA
  • 8 years

Available From: Available

    Candidate's About

    With several months of experience in Retail Analytics, Propensity Modeling, Customer Analytics, NLP, and more, I have executed data-driven solutions to enhance efficiency and accuracy in data processing. Proficient in creating supervised and unsupervised models, I design, develop, and deploy ML models in production. Skilled in cloud-based platforms like Microsoft Azure and VMware cloud, I manage data validation, EDA, predictive analytics, and provide insightful visualizations using Power BI and Shiny

    Work & Experience

    S

    Senior Data Scientist

    May 22 - PRESENT

    Technology

    • Icon
      Python
    • Icon
      SQL
    • Icon
      Power BI
    • Icon
      Git
    • Lead data science projects and provide technical guidance.
    • Develop and deploy advanced analytical models and algorithms.
    • Extract insights from complex data sets and provide actionable recommendations.
    • Collaborate with cross-functional teams to drive data-driven decision making.
    • Stay updated with the latest advancements in data science and contribute to knowl...
    Read More
    S

    Senior Data Scientist

    Nov 19 - May 22

    Technology

    • Icon
      Python
    • Icon
      SQL
    • Icon
      Power BI
    • Icon
      Git
    • Led data science projects and provided technical guidance to the team.
    • Developed and deployed advanced analytical models and algorithms.
    • Extracted insights from complex data sets and provided actionable recommendations to stakeholders.
    • Collaborated with cross-functional teams to drive data-driven decision making.
    • Stayed updated with the latest advancements i...
    Read More
    D

    Data Science Consultant

    Jul 18 - Nov 19

    Technology

    • Icon
      Python
    • Icon
      SQL
    • Icon
      Power BI
    • Icon
      Git
    • Provided data science consulting services to clients at Capgemini Technology Services India Ltd.
    • Assessed client requirements and developed tailored data science solutions.
    • Conducted data analysis, modeling, and predictive analytics to derive actionable insights.
    • Collaborated with clients to understand business challenges and develop data-driven strategies.
    • ...
    Read More
    S

    Senior Business Analyst/ Business Analyst

    Mar 14 - Jul 18

    Technology

    • Icon
      SQL
    • Icon
      Python
    • Icon
      Power BI
    • Icon
      Git
    • Analyzed business processes and gathered requirements as a Business Analyst | Senior Business Analyst at Tata Consultancy Services.
    • Conducted stakeholder interviews and facilitated workshops to elicit business needs and objectives.
    • Developed comprehensive business requirement documents and functional specifications.
    • Collaborated with cross-functional teams to transl...
    Read More

    Approval Velocity

    SENIOR DATA SCIENTIST

    • After extensive data cleaning/preprocessing, validation, and exploration, a first-cut model was developed for predicting the potential approval time of a quote when submitted.
    • Data from four sources was gathered and used to train the model, incorporating Cluster analysis and Trend analysis to determine the optimal target class.
    • Features were designed, and the best-fit models for each geography were identified using the Python package lazypredict.
    • Imbalances in the target variable were addressed through hybrid sampling and similar quote analysis based on Euclidean distance.
    • Further model improvement was achieved by identifying and removing Hard Instances using the pyhard package and tuning the XGBoost model parameters.
    • The model results and critical insights were presented to the business stakeholders.
    • Deployment code was developed, and the ML model was successfully deployed on the Deep Thought platform.


    • Python
    • SQL
    • Power BI
    • GIT
    • R, Shiny App
    • SAS
    View Details

    Propensity to Expand

    SENIOR DATA SCIENTIST

    • Conducted an analysis of repeated customers, recency, frequency, and monetary value to identify potential customers likely to expand their license purchases.
    • Analyzed and treated outliers, missing data, and junk data in the dataset extracted from the SQL database.
    • Established the target variable based on business inputs.
    • Utilized domain knowledge and statistics to refine the model iteratively using the Random Forest algorithm.
    • Shared critical insights and recommendations with the business based on the model output and data exploration.
    • Created a wireframe for real-time production and model recommendations.
    • Techniques employed included SMOTE (Synthetic Minority Over-sampling Technique) and Random Forest (RF) algorithm with K-fold validation.


    • Python
    • SQL
    • Power BI
    • GIT
    • R Shiny App
    • SAS
    View Details

    Customer Health Score

    SENIOR DATA SCIENTIST

    • Identified at-risk customers and prescribed call-to-actions based on an assessment of four categories/pillars: customer experience, business relationship, value & adoption, and product health metrics.
    • Gathered data from multiple sources, including SQL Server, survey data, and Redshift.
    • Assessed data validity and addressed potential challenges impacting overall customer health scores.
    • Created ten customer segments under each category using K-means clustering.
    • Derived customer health scores using multinomial logistic regression and aggregated scores from the four pillars to calculate a final health score.
    • Determined reasons for Customer Health Scores using input attributes.
    • Techniques utilized included K Means Clustering and Multinomial Logistic Regression.
    • Python
    • SQL
    • Power BI
    • GIT
    • R Shiny App
    • SAS
    View Details

    Software Dependency

    SENIOR DATA SCIENTIST

    • The Software Dependency tool identified software dependencies at a product-version level to support the sales team in ensuring customers have all necessary software systems.
    • Deduplicated customer product data using text similarity algorithms after gathering data from various sources.
    • Conducted detailed data exploration and analysis of customer data from multiple source systems, including coverage analysis and product data exploration.
    • Created ML models using APRIORI and SPM algorithms to determine product version dependencies.
    • Techniques employed included Sequential Pattern Mining (SPM), Apriori Algorithm, and Jaro-Winkler algorithm (Text Similarity algorithm).
    • Python
    • SQL
    • Power BI
    • GIT
    • R Shiny App
    • SAS
    View Details

    Code-Vox

    SENIOR DATA SCIENTIST

    • Code-Vox analyzed data from various sources (commit logs, code reviews, requirements management, static code analysis) to determine riskier components or source files in software development.
    • NLP techniques were employed to extract structured data from unstructured commit logs.
    • Developed the "Change Burst" metric to capture code quality metrics from structured data.
    • Classified files into high, medium, and low risk categories using a multi-layered ensemble model.
    • Provided recommendations for high-risk source-code files and components on a Shiny app and created an R package, 'Code Vox,' for automated processing.
    • Documented the package, its input, output, and capabilities, and developed a quality assessment checklist.
    • Utilized tools and techniques such as R, Shiny, GIT Bash, NLP, Gradient Boosting (GBM), Logistic Regression, KNN, and RF.
    • R Shiny
    • GIT
    • Bash
    • NLP
    • Gradient Boosting (GBM)
    • Logistic Regression
    • KNN
    • RF
    View Details

    Test-Vox / Test Selection Prioritization

    SENIOR DATA SCIENTIST

    • Test-Vox identified test cases that tended to fail together and prioritized them, reducing the need for retests and saving time and resources.
    • Data from tests, defects, requirements, project repositories, and code-vox recommendations was used in the prioritization and selection process.
    • The Apriori algorithm was utilized to discover association rules among the test cases.
    • A pattern was identified in test cases that frequently failed together, and the test case list was ranked for testing based on support and confidence in the association rules.
    • Tools and techniques employed included R, Shiny, and the Apriori algorithm.
    • R
    • R, Shiny
    • Apriori
    View Details

    New stores’ turnover Estimation

    SENIOR DATA SCIENTIST

    • Calculated turnover for a new store, facilitating the monitoring and identification of exceptional sales.
    • Cleansed, preprocessed, and extracted relevant data for analysis.
    • Established a baseline model using the averaging method for comparison with linear regression models.
    • Developed multiple regression models using SAS procedures.
    • Analyzed model assumptions and evaluated them using evaluation metrics like R-square, MAE, and RMSE.
    • Produced an analytical report and shared the results with the client.
    • Utilized tools and techniques such as SAS and Multivariate Linear Regression.


    • SAS
    • Multivariate Linear Regression
    View Details

    Sales Forecasting

    SENIOR DATA SCIENTIST

    • Forecasted and reported sales for a selected manufacturer for the next six weeks to aid in inventory planning based on trends, seasons, and cyclicity in the input data.
    • Implemented model adjustments and refreshes based on an analysis of forecast performance over time.
    • Provided insights into performance by analyzing performance metrics and competitive factors.
    • R
    • Shiny
    View Details

    Build Loss Prediction

    SENIOR DATA SCIENTIST

    • Predicted the potential number of vehicles to be built in the upcoming 12-18 months at different warehouses to assist in designing the optimal SIOP (Sales, Inventory, and Operations Planning) for manufacturing operations management.
    • Developed an ML model using historical SIOPs, builds, port delays, labor availability, machine availability, suppliers' data, inventory of plants, and external data.
    • Created an Alteryx pipeline to extract the correct data weekly and conducted data exploration, feature engineering, and coverage analysis.
    • Designed and developed an XG Boost regression model using Data Robot.
    • Produced a Power BI report for manufacturing operations management, analyzing builds, forecasts, product mix, and capacity utilization.
    • Data robot
    • Alteryx
    • Databricks
    • XGBoost
    • Power BI
    View Details

    Education

    B

    M.Sc. Statistics

    Bangalore University
    2011 - 2013
    V

    B.Sc. (MCS)

    Vijay College
    2008 - 2011

    Certificates