|
β½ Enterprise Operations 350+ Retail Outlets Monitored daily sales & stock at BPCL |
π Cloud Relational SQL 99K+ Orders Analyzed Target BigQuery multi-table case study |
π Database Competency 5-Star Gold Badge HackerRank verified SQL ranking |
π Academic Merit 8.67 CGPA B.E. Mechanical Engineering (VTU) |
I am a Data Analyst with a degree in Mechanical Engineering (8.67 CGPA) and a year of on-ground enterprise operations experience at Bharat Petroleum Corporation Limited (BPCL).
- βοΈ From Physical Systems to Data: My engineering background trained me to think in numbers, constraints, and root causes. At BPCL, I saw how physical operations rely on day-to-day data accuracy: monitoring fuel throughput, underground tank inventories, and dealer credit risk across 350+ retail stations.
- π οΈ Practical Process Automation: Working with SAP ECC, Advanced Excel, and VBA, I rebuilt manual daily reporting routines, cutting MIS compilation time from 2 hours down to around 30 minutes (~75% reduction) and evaluated bank credit eligibility under the e-DFS facility for 50+ retail dealers.
- π Rigorous Upskilling with Scaler DSML: I committed to deep analytical training through Scaler Academy's Data Science & Machine Learning program, mastering SQL (BigQuery, MySQL), Python (Pandas, NumPy, Scipy, Seaborn, Matplotlib, Sci-Kit Learn), Applied Probability, Hypothesis Testing, and Tableau Public. Learning Product analytics (RCA, RFM customer segmentation,Guess Estimates)
- π― Current Focus: Open to full-time Data Analyst / Business Analyst roles where I can combine operational intuition with clean SQL queries, statistical rigor, and executive dashboards.
| Languages & Querying |
|
| Analytics & Statistics |
|
| BI & Data Visualization |
|
| Enterprise & Workflow |
|
π Delhivery: Logistics Feature Engineering & ETA Optimization β Python SciPy scikit-learn
- π― Business Problem & Scale: Rebuilt 14,787 end-to-end trips from 144,867 raw checkpoint scans to test the predictive accuracy of Delhivery's Open Source Routing Machine (OSRM).
- π‘ Quantified Outcome: Actual trips ran 194.6 minutes longer than OSRM predictions on average (median 80 min,
$p < 0.001$ ), proving that OSRM underestimates real transit by failing to account for highway congestion and checkpoint handling. - π Direct Artifacts: π GitHub Repository β’ π Jupyter Notebook β’ π Case Study Report (PDF) β’ ποΈ Kaggle Dataset
π Target E-Commerce: BigQuery Logistics Disparity & Revenue Scaling β Google BigQuery SQL CTEs Window Functions
- π― Business Problem & Scale: Evaluated delivery lead-times across Brazilian states, transit bottlenecks, and customer repeat behavior across 99,441 orders in 6 relational BigQuery tables.
- π‘ Quantified Outcome: Identified a 3.5x regional delivery delay (29.3 days in Roraima vs. 8.3 days in SΓ£o Paulo), while JanβAug revenue grew +137% YoY. Recommended localized regional fulfillment hubs in the Northern zone.
- π Direct Artifacts: π GitHub Repository β’ π Master SQL Queries (
.sql) β’ π Case Study Report (PDF)
ποΈ Walmart: Consumer Purchase Behavior & Central Limit Theorem β Python CLT Sampling Distributions
- π― Business Problem & Scale: Tested demographic spending assumptions across 550,000+ Black Friday transactions to guide seasonal inventory allocation and promotional budgets.
- π‘ Quantified Outcome: Men spend +$702.96 (+8.05%) more per transaction with zero CI overlap at 99% confidence, while marital status showed no spending difference ($4.74 variance, completely overlapping CIs).
- π Direct Artifacts: π GitHub Repository β’ π Jupyter Notebook β’ π Case Study Report (PDF) β’ π± LinkedIn Slides
π² Yulu Micro-Mobility: Statistical Hypothesis Testing (ANOVA & Chi-Square) β Python Pingouin Two-Way ANOVA
- π― Business Problem & Scale: Diagnosed electric cycle revenue slumps and evaluated environmental and temporal drivers across 10,886 hourly rental records using a 6-step hypothesis testing framework.
- π‘ Quantified Outcome: Revealed a +136% weekend surge in casual riders against commuter-dominated weekdays, a 50% spring demand collapse (
$p < 10^{-100}$ ), and a 42% rental drop during adverse weather. - π Direct Artifacts: π GitHub Repository β’ π Jupyter Notebook β’ π Case Study Report (PDF)
π AeroFit: Treadmill Customer Segmentation & Conditional Probability β Python EDA Contingency Tables
- π― Business Problem & Scale: Evaluated customer demographic profiles across entry-level (KP281), mid-tier (KP481), and commercial (KP781) treadmills to build targeted sales recommendations.
- π‘ Quantified Outcome: Self-rated fitness was the single strongest indicator: 93.5% of customers rating fitness 5 bought the premium KP781 model, proving fitness habit outweighs income or age alone.
- π Direct Artifacts: π GitHub Repository β’ π Jupyter Notebook β’ π Case Study Report (PDF)
π Superstore: Executive Sales & Profitability Dashboard β Tableau Public LOD Expressions Parameters
- π― Business Problem & Scale: Provided executive visibility into loss-making product categories, regional discount sensitivity, and customer margins across $733K+ in total sales.
- π‘ Quantified Outcome: Discovered that aggressive discounting on Tables in the Central region accounted for over 60% of total regional operating losses, providing quantitative justification for discount caps.
- π Direct Artifacts: π Live Interactive Dashboard (Tableau Public) β’ π€ Tableau Public Profile
π View More Specialized Projects & Operational Analyses (BPCL, Rapido, HR, IoT, Netflix)
- β½ BPCL Retail Operations MIS (Excel, VBA, SAP ECC): Rebuilt manual routines for 350+ stations, cutting reporting time by 70% (2.5 hrs
$\rightarrow$ 25 mins) and assessed bank credit eligibility under e-DFS for 50+ retail dealers. - π Rapido Ride-Sharing Analytics (15 SQL Queries): Evaluated morning peak cancellation root-causes, window rankings, and distance-based user segmentation.
- π₯ HR Workforce Equity Analytics (9 SQL Queries): Analyzed departmental turnover risks, spans of control, and regional salary equity benchmarks.
- π Deloitte Australia Daikibo Factory IoT (Tableau): Analyzed industrial machine telemetry data (temperatures/vibrations) to isolate unscheduled downtime.
- π¬ Netflix Content Strategy EDA (Python & Seaborn): Analyzed international release trends and genre mix evolution across 8,807 titles.
- π HackerRank SQL - 5-Star Gold Badge (View Verified Profile)
- π¦πΊ Data Analytics Job Simulation β Deloitte Australia (via Forage) (Certificate)
- π SQL Skill Mastery Certification β Scaler Academy & InterviewBit
- π Data Analytics & Visualisation: Probability & Statistics β Scaler DSML
- π Data Analytics & Visualisation: Python Libraries (Pandas, NumPy, Seaborn) β Scaler DSML
- π Tableau & Excel Specialization β Scaler DSML
- π Bachelor of Engineering (B.E.) in Mechanical Engineering β VTU Belagavi (CGPA: 8.67)
"Numbers have an important story to tell. They rely on you to give them a clear and convincing voice."
β Stephen Few
Whether it's auditing daily sales across 350 petroleum stations, diagnosing regional logistics delays in BigQuery, or finding purchase drivers with conditional probability, I enjoy the craft of making data clean, clear, and actionable.