Analytics & BI
SQLPower BITableauLookerExcelKPI Reporting
Open to Data, BI, Product Analytics & AI Decision-System Opportunities
Swapnil
Herwadkar
Data & BI Analyst · Analytics Automation · AI-Powered Decision Systems
I turn ambiguous business problems into measurable analytics and decision systems — from SQL-backed reporting and executive dashboards to workflow automation, predictive analytics, and evidence-grounded AI products.
3+ yrsAnalytics & business intelligence experience
60+ hrsMonthly reporting effort saved through automation
30%Faster recurring report turnaround
25%Reduction in manual data processing effort
15%SQL query execution-time improvement
35%Workflow-efficiency improvement in reporting layers
01 / CAREER
Professional Experience
Operational analytics, enterprise reporting, business intelligence, and AI-assisted decision systems.
Nov 2025 — Present · Independent Analytics & AI ProjectsDecision Systems, Product Analytics & Sports AnalyticsIndependent
- Built SwapOpt, a local-first career decision system that compares job opportunities against verified candidate evidence and preserves hard rules such as missing evidence = UNKNOWN, scheduled interview ≠ occurred interview, accepted offer ≠ placement, and no autonomous application submission.
- Built the FPL Decision Engine as a local-first football decision-support system: Official FPL data and squad rules, LiveFPL rank/effective-ownership context, AI-assisted proposals and deterministic validation of transfer, budget, lineup and captaincy constraints. The public portfolio contains curated historical source and documentation, not the full later private runtime.
- Designed a human–AI research workflow around initial beliefs, model suggestions, manager choices and eventual outcomes to investigate decision quality, confidence and bias. Causal findings and improvements in FPL rank have not been established.
- Built the FIFA World Cup 2026 Sponsorship Analytics Platform with Python ETL, Snowflake, SQL, Power BI, Streamlit and OpenAI for sponsorship-market analysis and conversational analytics.
Oct 2024 — Oct 2025 · Data Analyst · ContractOperational Analytics, KPI Design & Reporting AutomationCommunity Dreams Foundation
- Translated homeowner intake, contractor matching, incentives, and project workflows into KPI definitions and dashboard requirements.
- Automated SQL-driven reporting and sustainability KPI workflows, reducing manual processing effort by 25%.
- Optimized MySQL reporting datasets and reusable BI logic for self-service analytics.
Nov 2022 — Jun 2023 · Data Analyst IIEnterprise BI, KPI Reporting & ETL AutomationAccenture
- Built SQL and Power BI KPI dashboards for recurring operational reporting.
- Automated ETL/reporting pipelines with scheduling, retries, alerts, and QA checks, saving 60+ hours per month.
- Standardized KPI definitions, mappings, validation rules, and reporting documentation.
Feb 2021 — Nov 2022 · Data Analyst ISQL Performance, ERP Data & Reporting ReliabilityAccenture
- Optimized high-volume SQL using indexing, partitioning, CTE refactors, and window functions, reducing execution time by 15%.
- Integrated SAP HANA/BW, ERP, and source-system data into reporting layers, improving workflow efficiency by 35%.
- Reconciled ERP, SQL, and reporting-layer data to improve accuracy and reduce discrepancies.
02 / SKILLS
Core Skills
Business-facing analytics with engineering depth for automation, validation, and scalable reporting.
Data Platforms
SnowflakeMySQLBigQueryRedshiftSAP HANASAP BW
Automation & AI
PythonPandasREST APIsNode.jsOpenAI APIETL
Business Analysis
RequirementsData ValidationRoot Cause AnalysisUATAgileJIRA
03 / PROJECTS
Selected Projects
From business intelligence and automation to AI-assisted decision systems and sports analytics. These projects show how I frame problems, work with evidence and turn ideas into practical tools.
Flagship · Evidence-Grounded AI Decision System
SwapOpt — AI Career Operating System
Local-first career decision system built around a different question: not “how fast can I apply?” but “does this opportunity deserve my effort?” It connects opportunity discovery, verified Candidate Identity, structured Role Identity, truthful application workflows, explicit interview/offer lifecycle, Network DNA, market intelligence, and outcome learning.
JavaScript · Node.js · Express · OpenAI · Chrome MV3 · GitHub Actions
Final private-runtime release: 3,244 / 3,244 automated tests · 24 modules · 257 validated paths · manual browser/runtime acceptance.
Build model: AI-assisted development with primary ownership of product requirements, workflow design, system logic, validation, debugging, and release execution.
View Public Showcase ↗Read Product Case Study ↗Explore Technical Blueprint ↗Sports Analytics · Explainable Decisions · Human–AI Research
FPL Decision Engine
The question: A player can be a great pick on paper without being the right transfer for your actual team. I built a local-first decision-support workflow to compare HOLD, transfers, captaincy and team choices using football evidence, available budget, existing players and the cost of alternatives.
Trusted data→Compare actions→Validate rules→Human decision
Official FPL is authoritative for squad state and scoring; LiveFPL supplies projected-rank and effective-ownership context. The AI proposes options, while deterministic checks assess squad, club, budget, lineup and captaincy constraints. Real transfers are never automated.
Research direction: How do AI suggestions affect a manager's initial judgment, confidence and choices? The evaluation design is public; improved FPL rank or causal effects are not yet demonstrated.
PowerShell · HTML / JavaScript · Official FPL API · LiveFPL · AI-assisted reasoning · Constraint checks
Public release: curated historical V4.1.2 source, architecture and a synthetic decision walkthrough. The newer private Windows runtime and personal FPL data are not published.
Published Writing
A Good Prediction Doesn't Always Make a Good Decision
AI + BI Platform
FIFA World Cup 2026 Sponsorship Analytics
End-to-end sponsorship analytics platform combining Python ETL, Snowflake, SQL, Power BI, Streamlit, and conversational AI for governed KPI reporting and decision support.
Python · Snowflake · SQL · Power BI · Streamlit · OpenAI
View Repository ↗Operations Analytics
ABS Inventory Forecasting & Optimization
Time-series inventory analysis using historical sales, replenishment, and lead-time data to forecast demand and support reorder-point, safety-stock, and SKU-allocation decisions.
Python · Pandas · Statsmodels · ARIMA · SARIMA
View Repository ↗Healthcare Supply Chain
ARV / HIV Treatment Distribution Analysis
Tableau-led analysis of treatment and test-kit shipments across Africa, examining delivery delays, vendor performance, shipment modes, pricing, and geographic distribution.
Tableau · Excel · XLMiner · Supply Chain Analytics
View Repository ↗Predictive Analytics
Ad Click Prediction
Predictive targeting pipeline for high-probability ad-click users using feature engineering, class balancing, and gradient-boosted modeling.
Python · Scikit-learn · XGBoost · SMOTE · Pandas
View Repository ↗ESG + Predictive Modeling
Fashion Sustainability Index
Analysis of fashion-industry environmental indicators with Random Forest modeling and Tableau dashboards to identify sustainability drivers and predict rating patterns.
Python · Scikit-learn · Random Forest · Tableau
View Repository ↗04 / EDUCATION
Education
Business analytics training built on an engineering foundation.
Sep 2023 — Aug 2024
Master of Science in Business Analytics
University of Massachusetts Amherst
Jul 2016 — Aug 2020
Bachelor of Engineering in Electronics
Ramdeobaba College of Engineering & Management
05 / CONTACT
Let’s Build Something Useful
Open to conversations around analytics, business intelligence, data products, sports analytics research, and AI-assisted decision systems.
Business Problems → Measurable Analytics → Better Decisions
My strongest work sits where business context, reporting, data quality, automation, and decision support intersect. SwapOpt is the clearest example of how I structure an ambiguous problem into an evidence-backed system.