Skip to content
View Yashdudhani's full-sized avatar
0 followers · 2 following

Achievements

Achievement: Pair Extraordinaire

Achievements

Achievement: Pair Extraordinaire

Block or report Yashdudhani

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Yashdudhani/README.md
Typing SVG

Overview

Education: B.E Computer Engineering at Dr. Babasaheb Ambedkar Technological University
Current Role: Data Analyst Intern at Pentagon Space
Profile: Building data-driven solutions that reduce decision time by 40%. From SQL query to Power BI dashboard to business insight — I own the complete analytics workflow.

Featured Projects

Three production-grade analytics systems built end-to-end — from raw data ingestion to executive dashboards — each designed to compress decision cycles and surface business risk before it costs money.

🏆 01 — Vendor-Insight-360 · End-to-End Vendor Analytics Platform

Compressed vendor evaluation from 2 weeks → 2 days across 200+ vendors and 200K+ records.

Key Achievements

  • 🔍 Engineered a leakage-safe churn classifier (ROC-AUC 0.73 vs 1.7% base rate) to proactively flag at-risk vendors before contract renewal windows
  • 📈 Built a backtested Holt-Winters demand forecaster achieving MAPE 0.67% — outperforming the naive baseline (0.87%) and enabling procurement teams to plan 60 days ahead
  • 🧩 Deployed K-Means vendor segmentation with silhouette-optimised k, turning 200+ unstructured vendor records into 4 actionable tier groups for negotiation strategy
  • ⚡ Automated the full reporting workflow — alert scheduling, PDF/Excel/HTML export, and JWT-authenticated REST API — eliminating 5+ hours of manual effort per week
  • 🔐 Shipped a 20-test Pytest suite with data-contract and leakage checks, ensuring production-grade data quality at every pipeline stage

Tech Stack

Python SQL Streamlit scikit-learn Plotly Flask SQLite Pytest

Impact Metrics

Metric Result
Vendor Eval Time 2 weeks → 2 days
Records Processed 200K+
Forecast MAPE 0.67%
Manual Hours Saved 5+ hrs/week
Test Coverage 20 automated tests

View Code   Pentagon Space

💳 02 — HSBC Credit Card Intelligence Dashboard · Enterprise BI & Risk Analytics

Transformed transactional credit card data into a real-time executive command centre for spend, risk, and customer segmentation.

Key Achievements

  • 📊 Designed a multi-page Power BI dashboard tracking spend velocity, credit utilisation, and delinquency trends across customer cohorts — enabling risk teams to act on signals 3× faster
  • 🧮 Wrote complex DAX measures and calculated columns to model rolling 90-day delinquency rates, utilisation buckets, and month-over-month spend variance at the card-holder level
  • 🗄️ Built and optimised the SQL data model (star schema) powering sub-3-second dashboard refresh across millions of transaction rows
  • 👥 Applied RFM segmentation (Recency, Frequency, Monetary) to stratify 50K+ cardholders into 5 behavioural tiers, directly informing retention and upsell campaigns
  • 🛡️ Embedded row-level security (RLS) in Power BI to enforce data access governance across business units — a production requirement missed by most BI portfolios

Tech Stack

Power BI DAX SQL Python Pandas Excel

Impact Metrics

Metric Result
Dashboard Refresh < 3 seconds
Cardholders Segmented 50K+
Risk Signal Lead Time 3× faster
Schema Design Star schema, RLS-enforced
Certification Aligned PL-300 (Jun 2026)

View Code   Microsoft Certified

🛍️ 03 — Shopeasy Consumer Intelligence Funnel & Sentiment Analysis · Behavioural Analytics + NLP

Mapped the full consumer journey from first click to churn signal — combining funnel analytics with NLP-driven sentiment scoring to pinpoint exactly where and why customers drop off.

Key Achievements

  • 🔁 Reconstructed a multi-stage conversion funnel (Awareness → Consideration → Purchase → Retention) from raw event logs using SQL window functions, exposing a 34% drop-off at the cart stage invisible to the existing reporting stack
  • 💬 Applied NLP sentiment analysis (VADER + custom lexicon) across 10K+ product reviews to produce a per-category sentiment score, correlated against return rates to identify product-quality risk signals
  • 📉 Built a cohort retention heatmap in Python (Seaborn + Matplotlib) segmented by acquisition channel, revealing that paid-social cohorts churned 2.1× faster than organic cohorts within 30 days
  • 📊 Delivered a Streamlit executive dashboard with filterable KPIs — conversion rate, average order value, NPS proxy score, and sentiment trend — enabling non-technical stakeholders to self-serve insights
  • 🔬 Conducted A/B test significance analysis (chi-squared + effect size) on two checkout UX variants, providing a statistically grounded recommendation that informed the product roadmap

Tech Stack

Python SQL Streamlit NLTK Pandas Seaborn Matplotlib SciPy

Impact Metrics

Metric Result
Drop-off Identified 34% at cart stage
Reviews Analysed 10K+
Churn Insight Paid-social 2.1× faster
A/B Test Rigor Chi-squared + effect size
Stakeholder Tool Self-serve Streamlit app

View Code

📌 What These Projects Prove

Capability Evidence
🗄️ Advanced SQL Star schema design · window functions · complex aggregations · sub-3s query performance
📊 Power BI Multi-page dashboards · DAX · RLS governance · exec-ready storytelling
🐍 Python for Analytics EDA pipelines · ML models · NLP · statistical testing · automated reporting
📈 Business Impact Every project ships a measurable outcome — not just a model, but a decision
🏢 Enterprise Readiness JWT auth · CI/CD · data contracts · row-level security · Pytest suites

Certifications & Credentials

Certification Issuer Date Credential
Microsoft Certified: Power BI Data Analyst Associate (PL-300) Microsoft Jun 2026 Verify Credential
IBM Data Analyst Professional Certificate IBM Coursera Dec 2025 Verify Badge
Deloitte Data Analytics Job Simulation Forage Jun 2025 Verify Certificate
Published Researcher IJARSCT Journal Jun 2025 Read Paper

Skills Validated:

  • ✅ Data Analysis & Statistical Insights
  • ✅ SQL & Database Management
  • ✅ Python for Data Processing
  • ✅ Tableau & Power BI Dashboard Design
  • ✅ Exploratory Data Analysis (EDA)

GitHub Stats

📊 My GitHub Analytics

Metric Info
📁 Public Repositories View on GitHub
💻 Primary Languages Python, SQL
🔧 Tech Stack Pandas, NumPy, Power BI, Streamlit
📈 Main Focus Data Analysis & Analytics
📚 GitHub Profile @helloworld880

🎯 Key Contributions

  • Vendor-Insight-360 — Production Analytics Platform Built a modular Streamlit analytics platform tracking 120 vendors across a 24-month dataset. Engineered a leakage-safe churn classifier (ROC-AUC 0.73 vs 1.7% base rate), a backtested Holt-Winters forecaster (MAPE 0.67% beating a naive 0.87% baseline), K-Means vendor segmentation with silhouette-selected k, and hypothesis tests (Welch's t / ANOVA / chi-squared) reported with effect sizes. Includes a JWT-authenticated Flask REST API, PBKDF2 login, automated alert and report scheduler, PDF/Excel/HTML export, and a 20-test Pytest suite with data-contract and leakage checks. Stack: Python · Streamlit · SQLite · scikit-learn · statsmodels · SciPy · Plotly · Flask · Pytest

  • Invoice Payment Intelligence — Enterprise AI Risk Platform Developed an invoice payment-delay prediction system combining a traditional ML engine (Random Forest / XGBoost, 87% accuracy) with a deep learning engine (TensorFlow/Keras LSTM, 92% accuracy). Integrated Apache Spark for batch processing at 15,000+ invoices/second, PostgreSQL for persistence, and Docker + docker-compose for deployment. Set up GitHub Actions CI/CD pipelines and a multi-tier test suite (unit, integration, performance benchmarks). Stack: Python · TensorFlow · scikit-learn · Apache Spark · PostgreSQL · Docker · GitHub Actions

  • House Price Prediction — Published Research + Deployed ML App Conducted end-to-end EDA and predictive modelling on 14,619 Indian real estate records (23 features). Trained and serialized a regression model, deployed it as a Streamlit web app for live price estimates, and published findings as a peer-reviewed paper (DOI: 10.48175/IJARSCT-27446, IJARSCT June 2025). Stack: Python · scikit-learn · Pandas · Streamlit · Jupyter

Visit My GitHub Profile →


Connect With Me

🌐 Social & Professional Links

LinkedIn Gmail GitHub Portfolio


📝 Latest Work & Publications

🎯 Vendor-Insight-360

End-to-End Analytics Platform for Vendor Performance Management

🏢 Organization: Pentagon Space, Bengaluru ⏱️ Duration: Mar 2026 - Present (3 months) 🛠️ Tech Stack: Python, SQL, Pandas, NumPy, Tableau, Power BI 📈 Impact: Reduced vendor evaluation time from 2 weeks to 2 days 📊 Scope: 200+ vendors tracked, 200K+ records processed 🔧 Deliverables: Automated data pipeline + Interactive dashboards + SQL analytics

💻 View Code

📚 Real Estate Price Prediction

Published in International Journal of Advanced Research in Science, Communication and Technology (IJARSCT)

📖 Title: Real Estate Price Prediction Using Data Analysis 📅 Date: June 2025 | Volume 5, Issue 2 🔗 DOI: 10.48175/IJARSCT-27446 👥 Team: Yash Dudhani, Dastagir Sutar, Adil Baig, Samarth Hatture 📊 Highlights: Complete EDA, statistical analysis, data-driven insights

📖 Read Full Paper | 🔍 Verify on Crossref


What I'm Working On

  • 📊 Advanced SQL optimization & complex queries
  • 🎨 Advanced data visualization & storytelling techniques
  • 📈 Statistical analysis for business insights
  • ☁️ Cloud-based analytics platforms
  • 📚 Publishing more research & case studies

Open to Opportunities

💼 Roles: Data Analyst | Junior Data Analyst | Business Analyst | BI Analyst

🌍 Locations: Bengaluru | Hyderabad | Pune | Remote-friendly

🏢 Employment Types: Full-time | Internship | Contract | Advisory

📧 Get in Touch: Email | LinkedIn



⭐ If you found this README helpful, consider giving it a star!
Made with ❤️ by Yash Dudhani

Popular repositories Loading

  1. House-Price-Prediction-Python House-Price-Prediction-Python Public

    Jupyter Notebook

  2. invoice_payment_Python invoice_payment_Python Public

    Python

  3. Vendor-Insight-360 Vendor-Insight-360 Public

    Python

  4. Shopeasy-consumer-intelligence-funnel-and-sentiment-analysis Shopeasy-consumer-intelligence-funnel-and-sentiment-analysis Public

    Jupyter Notebook

  5. Credit-Card-Intelligence-Dashboard Credit-Card-Intelligence-Dashboard Public

    Jupyter Notebook

  6. Yashdudhani Yashdudhani Public