Hello!

I'm Harshal Kamble



A Data Scientist and Analyst

My works

Name: Harshal Kamble

Job Role: Data Analyst

Experience: 3 Year

Address: Atlanta, USA

Skills

SQL 95%
PYTHON 95%
Data Visualization 100%
Statistical Analysis 95%
Machine Learning 90%

About

About Me

I'm a Data Analyst currently at Amazon, working on building SQL pipelines, automating reporting workflows, and partnering with cross-functional teams to turn data into decisions. Before this, I worked as a Graduate Research Assistant at Georgia State University on projects spanning healthcare analytics, generative AI, and predictive modeling. I love working with data — whether that's writing complex SQL queries, building dashboards in Power BI and Tableau, or experimenting with ML models in Python. Outside of work, I'm always exploring new tools, side projects, and ways to keep learning in the data space.

  • Profile: Data Science & Analytics, AI/ML
  • Education: Masters in Data Science
  • Programming: Python, R-Programing, SQL
  • DataBase : MySQL, PostgreSQL, SQL Server, MongoDB, AWS
  • Tools and FrameWorks : Tableau, Power BI, Matplotlib, Seaborn, Excel, Databricks
  • Machine Learning: Isolation Forests, Decision Trees, Regression, Clustering, Time Series, NLP.
  • Language: English, Japanese, Hindi & Marathi
  • Certifications: Microsoft Certified: (PL 300) Power BI Data Analyst Associate, Python (University of Michigan), Introduction to Statistics (Stanford University)

0 +   Projects completed

LinkedIn

Resume

Resume

Seasoned Data Analyst with 3 year of experience driving business strategies through data-driven insights. Proven expertise in data science, statistical analysis, machine learning algorithms and project management.

Download Resume

Experience


March 2026 - Present

Amazon

Business Analyst II
  • Re-engineered a legacy SQL analytics pipeline using recursive CTEs, window functions, and query optimization on Amazon Redshift cutting runtime by 60% and unlocking ~$800K in projected annualized savings.
  • Automated end-to-end data ingestion (S3 → Redshift → Python ETL → Tableau/QuickSight), saving ~10 hours per week of analyst effort and standardizing intake across operations and finance.
  • Partnered with Product, Ops, and Finance leaders to standardize 8+ core business metrics, unifying executive reporting and eliminating conflicting numbers across teams.
  • Reduced insight-delivery turnaround from 2 days to under 1 hour through reusable SQL/Python templates and self-serve QuickSight dashboards, enabling 50+ stakeholders to answer ad-hoc questions independently.
Aug 2025 - Dec 2025

Georgia State University

Graduate Research Assistant – Data Scientist
  • Cleaned and merged 3,000+ ABI patient NOMS and encounter records (Stroke & TBI) using Python and SQL, improving data accuracy to 98% for reliable rehabilitation analysis.
  • Developed interactive dashboards visualizing therapy intensity, payer mix, and recovery outcomes, enhancing clinician insight clarity by 40% and informing weekly review discussions.
  • Built Random Forest models on Self-Care and Mobility data to predict patient improvement, achieving R² = 0.81 and enabling therapists to identify high-risk patients 2 weeks earlier.
May 2025 - Aug 2025

Amazon

Business Intelligence Engineer
  • Redesigned 1,000+ lines of SQL workflows (recursive CTEs, schema remapping, window functions), reducing runtime by 40% and enabling scalable workforce analytics across 150+ sites.
  • Designed and optimized ETL workflows in SQL + Python on Amazon Redshift within the AWS ecosystem (ingest → transform → validate), adding schema checks that reduced refresh failures and lifted data reliability.
  • Automated BI dashboards by shifting from once-daily static refreshes to 4x daily dynamic updates, giving leaders at 150+ sites near real-time workforce visibility and cutting decision lag by 80%.
  • Partnered with cross-functional teams in Agile sprints to implement a standardized reporting framework, reducing reporting errors by 30% and strengthening trust in workforce planning data.
  • Designed and executed A/B tests on refresh strategies (1× vs 4× daily), using t-tests and confidence intervals to validate decision-lag reductions and staffing accuracy gains before global rollout.
  • Integrated AI forecasting models (time-series, pattern recognition) into dashboards, providing 2-week early warnings on staffing risks and informing staffing and budget decisions.
August 2024 - April 2025

Georgia State University

Graduate Research Assistant – Data Scientist
  • Consolidated 5+ clinical data sources into a unified Digital Twin dataset, improving data quality by 40% to support predictive simulations for treatment outcomes.
  • Designed a cognitive fatigue prediction model (MFIS scores) achieving 60% cross-validation accuracy, improving by 35% after bias correction to optimize treatment resource allocation.
  • Led development of a Generative AI solution on complaint data, boosting semantic accuracy from 50% → 81% and improving AI-generated response accuracy from 36% → 77% through prompt engineering and A/B testing.
  • Automated complaint classification and routing through AI-driven workflows, reducing manual interventions and cutting resolution time by 35%.
  • Built Power BI dashboards for real-time simulations and ran A/B tests (t-tests, chi-square, stratified sampling) to validate model improvements before rollout, reducing testing time by 30%.
July 2023 - July 2024

TDK Electronics

Data Scientist

TDK Corporation is a leading electronics company based in Tokyo, Japan. It was established in 1935 to commercialize ferrite, a key material in electronic and magnetic products.

  • Conducted product growth and profitability analytics across 80+ business units, improving identification of high-growth opportunities by 15% and enhancing resource allocation efficiency by 20%.
  • Integrated Power BI with Amazon Redshift using Row-Level Security (RLS), ensuring secure and compliant reporting access.
  • Assembled a Power BI model fed by scheduled SQL/Python ETL transformations; standardized dimensional tables and measures (DAX) to create a single source of truth for unit economics.
  • Benchmarked ARIMA vs. Random Forest forecasts with 80% accuracy, simulating A/B testing scenarios and applying hypothesis tests to guide procurement and pricing assumptions.



Education


2024-2025 (Pursuing)

Georgia State University

Master's in Data Science and Analysis

Grade: 3.8/4.0 GPA

2023-2024

MIT School Of Distance Learning

Postgraduate Certificate in Business Analyst

Grade: 4.0/4.0 GPA

2020-2023

Pune University

Bachelor in Computer Science

3.8/4.0 GPA

Projects

Projects

Below are the sample Data Analytics projects on SQL, Python, Power BI & ML.

ChatDesk AI – ChatGPT Desktop App

Python (Tkinter + CustomTkinter) desktop client for ChatGPT with model selection, session history, and optional text-to-speech using OpenAI’s Chat Completions API.

Official Georgia State University ChatBot

AI-powered chatbot that provides students with detailed course information without human intervention.

🛡️ Credit Card Fraud Detection

Real-time ML pipeline to detect fraudulent credit card transactions on a highly imbalanced dataset.

Image Based Deep Learning & Reinforcement Learning Strategy Design

Two-part project: image-based deep learning and RL strategy design.

Super Store Data Analysis using SQL

Advanced SQL analysis to identify gaps and drive business growth.

Data Analysis using Python Project

EDA on Diwali sales data to improve customer experience.

Random Forest – Time Series Forecasting

Predicting tokens using current and previous three months of data.

Analysis on Chick-fil-A

Extracting information from franchise documents using NLP.

Power BI Dashboard on Shipping Data

Interactive dashboard for 2024 shipping data across carriers, methods, regions, and pack types.

Apple's Sales Analysis in Power BI

Comprehensive analysis of Apple Inc. sales data in Power BI.

Apple Revenue Analysis Project

Trends and insights from Apple stock data related to revenue performance.

0 Projects
0 Cups of coffee

More projects on Github

I love to solve business problems & uncover hidden data stories


GitHub

Contact

Contact Me

Below are the details to reach out to me!

Address

Atlanta, USA

Contact Number

+1 (404)-723-6762

Email Address

harshalkamble2511@gmail.com

Download Resume

Resumelink




  • Find me on