Machine Learning Business Analyst

Adarsh Prajapat

Boston, MA 617-***-**** *******@**.*** www.linkedin.com/in/adarsh-prajapat https://github.com/Adarsh-1406

SUMMARY

Data-driven professional with 2+ years of experience in analytics, machine learning, and cloud-based pipeline development. Proven track record at Amazon in designing scalable dashboards using QuickSight and optimizing data processes to enhance decision-making. Proficient in SQL, advanced Excel, and predictive modeling to derive actionable insights and drive strategic impact. Seeking a Data / Business Analyst role to drive strategic impact through data-driven solutions.

EDUCATION

M.S. in Business Analytics (STEM) Aug 2023 - Jan 2025

Boston University, Questrom School of Business, Boston, MA

Bachelor of Technology - Electronics & Communication Engineering (ECE) Aug 2016 - Jul 2020

MCKV Institute of Engineering, Kolkata, West Bengal

SKILLS

•Programming: Python (NumPy, Pandas, Scikit-Learn, Matplotlib, Seaborn), SQL (PLSQL, MySQL, BigQuery), JavaScript

•BI / ETL Tools: Tableau, Amazon QuickSight, Power BI, MS Excel, Streamlit, Apache Superset, Prefect, Git, GitHub, JIRA

•Databases: MS SQL Server, AWS (Redshift, S3), Snowflake, BigQuery, MotherDuck, Pinecone

•Machine Learning: Supervised Learning (Random Forest, SVM, XGBoost, KNN), Unsupervised Learning (K-Means, PCA)

•Cloud Platforms: Google Cloud Platform (GCP) Services (Vertex AI, Compute Engine, Cloud Run Functions), AWS

•Statistics: Hypothesis Testing, A/B Testing, Regression, ANOVA, Time Series Analysis (ARIMA, SARIMA, SARIMAX)

•Tech Skills: Data Analysis, Data Visualization & Reporting, Data Modeling, Deploying Analytics Pipeline, Machine Learning

•Certifications: AWS Certified Cloud Practitioner (Ongoing), AWS Certified Data Engineer – Associate (Ongoing)

•Other Relevant Coursework: Financial and Accounting Analytics, Management Communication & Teaming

WORK EXPERIENCE

Analyst, Strategic Growth Business (BFSI), Tata Consultancy Services Mar 2021 – July 2023

•Contributed to the development and enhancement of web application and database using JavaScript and PL/SQL for banking and insurance clients within an Agile framework, driving efficiency and innovation.

•Facilitated seamless client interactions by gathering requirements through one-on-one calls, translating them into technical solutions, and delivering 100% quality outputs, leading to a measurable increase in client satisfaction.

•Collaborated across cross-functional, onshore and offshore teams to enhance project transparency and efficiency while resolving application defects in frontend and backend systems through in-depth root cause analysis, detailed documentation and rigorous unit testing, reducing project turnaround time by 10%.

Business Analyst (L4) Intern, Amazon Development Center India Pvt. Ltd. Feb 2020 - Jul 2020

•Migrated and optimized Tableau dashboards to Amazon QuickSight, utilizing advanced Excel functions (pivot tables, VLOOKUP, data validation), SQL for dashboard customization to improve usability, ensuring 100% data integrity.

•Helped drive down the cost of Fulfillment Centers (FCs) in respect of overall Tableau user licensing cost by 90% and improving dashboard efficiency for Fulfillment Centers (FCs) across North America.

•Developed QuickSight dashboards for the Reliability Maintenance and Engineering (RME) department, increasing user accessibility by 170% and enabling data-driven decision-making.

•Conducted a deep-dive root cause analysis on data anomalies, ultimately helping stakeholders and FCs address key operational inefficiencies, optimize maintenance processes, and reduce downtime.

PROJECT EXPERIENCE

Building Cloud Native Data Pipeline for BOSTON 311 Service Requests Sep 2024 – Dec 2024

Deploying Analytics Pipeline, Boston University

•Developed a cloud-native data pipeline for real-time ingestion, transformation, and storage of Boston 311 service requests of 100,000+ daily records using GCP, MotherDuck and Prefect, ensuring seamless data processing and dynamic updates.

•Built and deployed a predictive machine learning model to forecast Case Resolution Times, achieving 87% accuracy with Random Forest Regression (R2=0.87), reducing SLA non-compliance rates.

•Enhanced geospatial and interactive analytics by integrating Mapbox and Streamlit dashboards, providing stakeholders with location-based insights and real-time visualizations to prioritize high-risk cases and allocate resources efficiently.

•Implemented Text-to-SQL LLM capabilities enabling natural language querying of datasets, reducing query time by 30%, empowering non-technical users to generate reports and access insights, streamlining decision-making processes.

Analyzing Factors Behind Success in Modern TV Industry Nov 2023 - Dec 2023

Business Analytics Toolbox, Boston University

Spearheaded the team to create Entity Relationship Diagram for IMDb dataset, comprising 7 tables with data on 37M titles, enabling comprehensive analysis of TV industry data.

Used SQL in BigQuery to join and aggregate tables, uncovering key insights like top TV-show producing countries, trends in average runtime, and season lengths over time.

Managed the development of a Tableau dashboard that visualized key findings through interactive charts, resulting in improved data accessibility and decision-making for stakeholders.

Analyzing Hospital Readmission Rates for Diabetic Patients Aug 2023 – Oct 2023

Introduction to Data Analytics, Boston University

•Designed and trained machine learning models using Python to predict hospital readmission rates in diabetic patients, through in-depth analysis of demographic and medical history data, achieving a 15% model accuracy improvement.

•Addressed dataset inconsistencies and limited medical knowledge by feature engineering and optimization techniques, reducing diabetic patient readmission risk rate by 20%.

•Identified key predictors of hospital readmission, such as number of lab procedures, medications, and Length Of hospital Stay (LOS), leading to action-oriented insights for healthcare improvements despite model recall challenges.