Chethana Kuppala

01ABOUT

A blend of curiosity and proof.

Where discipline meets data — every dashboard I ship is backed by a pipeline I've personally run, broken, and fixed.

Chethana Kuppala is a business intelligence engineer with dual master's degrees in Business Analytics and Medical Informatics from the University of South Carolina, specializing in Power BI, DAX, and end-to-end SQL/Python ETL across AWS S3 and Snowflake.

She's built and optimized data warehouses, automated reporting pipelines, and CI/CD-gated quality controls that have cut reporting time by up to 55% and improved data-load accuracy to 98.95% — work she verifies end-to-end before it ever reaches a stakeholder deck.

Currently open to Business Analyst, Business Intelligence, and Data Engineer roles.

02EXPERIENCE

Where I've shipped this, in production and in class.

Aug 2025 —
May 2026

Power BI Developer — Fan & Ticketing Analytics

USC Athletics

Contributed to a 20% ticket-revenue lift — validated through variance analysis presented to 2 executive stakeholder groups — by building 7+ Power BI and Tableau dashboards with DAX-driven KPIs across 1M+ rows of sales and attendance data.

SQLPOWER BIDAXFORECASTING

Jan 2025 —
May 2025

Graduate Assistant — Analytics

Molinaroli College of Engineering & Computing

Cut reporting time by 55% by building executive-ready Power BI dashboards for 2 departments, and uncovered a 30% cost trend within a $15K–$20K budget by analyzing financial and utilization data in SQL, SAS, and Python.

SQLSASPYTHONPOWER BI

Aug 2025 —
May 2026

Graduate Assistant — Power BI / Data Analyst

Student Success Center

Improved data accuracy by 15% by developing and troubleshooting SQL queries and ETL-style report datasets, then root-causing discrepancies across multi-source data — plus automating workflows to cut manual reporting effort by 10+ hours/month.

SQLETLDATA VALIDATION

03PROJECTS

What I've actually shipped and run.

FEATURED · VERIFIED AGAINST LIVE AWS + SNOWFLAKE

Fintech Lease-to-Own Data Platform

An end-to-end ETL pipeline moving synthetic lease-to-own records from an AWS S3 landing zone through Python/SQL validation into a Snowflake warehouse (RAW → STAGING → ANALYTICS), gated by real CI/CD — not just designed to support it.

3,672 defects quarantined across 8 categories

24 tests passing on every push

AWS S3SNOWFLAKEPYTHON GITHUB ACTIONS CI/CDPOWER BI

MARKET RESEARCH · SEGMENTATION & TURF

Consumer Brand Health & Commercial Diligence — AG1

A synthetic 280-respondent brand health study of AG1 in the daily foundational nutrition category — rim-weighted funnel and NPS decomposition, k-means attitudinal segmentation, TURF line-extension optimization, Van Westendorp pricing, and three alternative-data sources triangulated into an 8-slide commercial diligence readout. 32 of 32 headline figures reconcile between an independent pandas engine and a DuckDB SQL layer within 0.15 points.

PYTHONSQLEXCEL

POWER BI · STAR-SCHEMA MODELING

PulseComply — LMS Training Compliance Analytics

A Power BI-ready star-schema model (fact/dimension tables, DAX measures) on a 1,400-employee dataset. Diagnosed a 27-point on-time compliance gap and a 68% first-attempt pass rate pointing to a quiz-design defect — then quantified a campaign that recovered 72.5% of a 661-item overdue backlog.

POWER BIDAXSTAR SCHEMA

EXCEL MODELING · VARIANCE ANALYSIS

Manufacturing Landed Cost & Supplier Variance Model

A standard-vs-actual landed cost model for a simulated manufacturer's FY2025 book — 390 transactions across 20 parts and 10 suppliers in 6 countries, decomposed into 6 variance drivers plus a mid-year tariff step-up study. $567,926 unfavorable variance found; $429,433 re-sourcing savings identified. 11,600+ live Excel formulas, zero recalculation errors.

EXCELPYTHONVARIANCE ANALYSIS

EXCEL MODELING · FUND ACCOUNTING

Cascade Capital — PE Fund Model

A full fund-accounting model for a simulated $50M PE fund — capital calls, a four-tier distribution waterfall, quarterly NAV rollforwards, and a GP-ledger-vs-fund-admin reconciliation across 15 LPs and 10 portfolio companies. 20.9% net IRR / 1.69x TVPI; all 15 LP capital accounts tie to fund NAV to the cent.

EXCELPYTHONIRR / MOIC

PROCESS ANALYSIS · AUTOMATION IMPACT

FlowForward — Process Blueprint

A current-state → future-state redesign of an indirect-procurement requisition workflow — swimlane process maps, root-cause analysis, and a tool/process/people redesign, with cycle-time and cost impact modeled by a calculation engine rather than hand-typed numbers.

PROCESS DESIGNPYTHON

EXCEL MODELING · CATEGORY STRATEGY

Supplier Portfolio Segmentation Model

Segments a 25-supplier portfolio on a Kraljic-style spend-vs-risk matrix and turns it into a ranked negotiation priority list. $4.48M in spend analyzed; $235,171 (5.2%) in negotiation savings identified across 7 strategic and 8 leverage suppliers.

EXCELPYTHON

EXCEL MODELING · PRICE VARIANCE

Should-Cost & Price Variance Model

Quantifies cross-location price variance on 225 purchase events across 4 sites and 20 items. $556,803 in modeled spend; $49,151 (8.8%) in savings identified if every site paid the best confirmed unit price.

EXCELPYTHON

EXCEL MODELING · SPEND ANALYTICS

Catalog Adoption & Maverick Spend Tracker

Quantifies catalog-adoption gaps across 480 simulated purchase orders. Found 60.7% of modeled spend was off-catalog — weakest in Professional Services (81% maverick), strongest in Office Supplies (8%) — and ranked the top 10 onboarding candidates.

EXCELPYTHON

SQL · BIGQUERY · A/B TESTING

GOTV Campaign Measurement

An end-to-end measurement framework for a 24,000-contact multi-channel voter outreach program — conditional funnel analysis, an A/B message test (two-proportion z-test, p < 0.001), send-time optimization, and voter file matching, in BigQuery SQL with Looker-Studio-style reporting.

SQLBIGQUERYPYTHON

AI AGENTS · STRUCTURED OUTPUT

ClarityBridge

Two AI business-analyst agents — one surfaces contradictions and risk in a messy request, one turns a vague problem into objectives, KPIs, and user stories — built on Streamlit, FastAPI, and Pydantic-validated structured LLM output, unit-tested with the API call mocked out.

PYTHONOPENAI APIFASTAPI

No projects match that filter yet — try another skill, or view everything on GitHub.

Every pipeline startsas noise. I ship it as signal.

A blend of curiosity and proof.

Where I've shipped this, in production and in class.

Power BI Developer — Fan & Ticketing Analytics

Graduate Assistant — Analytics

Graduate Assistant — Power BI / Data Analyst

What I've actually shipped and run.

Fintech Lease-to-Own Data Platform

PulseComply — LMS Training Compliance Analytics

Every pipeline starts
as noise. I ship it as signal.