Prajakta Shewale

Data Science | Machine Learning | NLP | Generative AI

Pune City

Summary

Results-Driven Data Scientist & AI Specialist with 3+ years of experience in building scalable Machine Learning, Generative AI, and Advanced Analytics solutions. Proficient in Python, SQL, and Power BI with expertise in NLP, predictive modeling, and AI-driven solutions. Skilled in building scalable models, developing intelligent applications, and transforming complex data into actionable insights to drive data-driven business decisions.

Overview

years of professional experience

Certifications

Work History

Assistant Manager (Data Science)

WNS

12.2024 - Current

Project: Insurance Data Reporting and Performance Analysis.
Tools: Python, SQL, Machine Learning, Natural Language Processing, Power BI, AWS (S3, Athena), Excel.
Migrated 20+ Excel reports to Power BI dashboards, reducing reporting time by 40%, and enabling real-time tracking of 50+ KPIs.
Integrated ML models (Random Forest, XGBoost) to deliver predictive insights for claim trends and risk scoring, improving accuracy by 15%.
Optimized SQL queries and managed transactional data pipelines, reducing execution time by 30%.
Applied NLP techniques to analyze customer-advocate conversations, deriving insights on customer satisfaction and behavior.
Implemented clustering models to segment customers and identify key drivers for business outcomes.

Software Engineer (Data Science)

Xoriant Solutions Pvt Ltd

08.2021 - 07.2023

Project 1: Predictive Modeling and Key Driver Identification Using Multi-Source Data.
Tools: Python, SQL, Machine Learning, Natural Language Processing, Visualizations, Streamlit.
Developed an end-to-end machine learning pipeline on multi-source structured datasets.
Performed large-scale EDA using Python, and built visualizations to uncover trends, improving analytical efficiency by 25%.
Built and optimized machine learning models, including preprocessing, feature engineering, and hyperparameter tuning, achieving approximately 15–20% improvement in prediction accuracy.
Applied unsupervised learning techniques on high-dimensional datasets to identify key patterns, and segment data effectively.
Engineered and high-impact features, identifying key drivers, and enabling a 15% improvement in model-driven outcomes and decision-making.
Project 2 : Intelligent Chat Analysis using NLP
Tools: Python, Machine Learning, Natural Language Processing, and Gen AI.
Leveraged advanced NLP techniques to analyze and classify chatbot interaction data (10K+ records), predicting purchase intent and satisfaction levels.
Implemented transformer-based models (BERT) to extract patterns from text data, improving classification accuracy by 25%.
Designed and deployed semantic similarity pipelines using BERT embeddings and cosine similarity, increasing intent classification accuracy by 15%.
Developed a BERT-based question answering system, improving response relevance and reducing query resolution time by 20%.

Education

Bachelor of Engineering (BE) - Information Technology

Savitribai Phule Pune University

Pune, India

01-2021

Skills

Language & Libraries : Python (Pandas , Numpy , Scikit Learn, NLTK, Spacy, Gensim) ,SQL
Visualization & BI : Power BI ,Seaborn , matplotlib
Machine Learning : Regression, Classification, Clustering, Feature Engineering, Time Series

Natural Language Processing : Text cleaning, Text Preprocessing, Tokenization, Embedding, Transformer , NER,Sentiment Analysis,Topic modeling
Cloud Platform : Amazon Web Services (AWS), Google Cloud,Azure
Statistics : Descriptive , Inferential

Websites

https://www.linkedin.com/in/prajakta-s-4329631b0ps21905/

Certification

Full Stack Data Science & AI – NASSCOM

Timeline