Computer Science × Economics · Purdue University · Class of 2027

Nitya
Jhaveri

I turn messy data into decisions.

Business analysis, data science, and economic modeling — with a bias toward problems where technology meets strategy. I've forecasted the U.S. alcohol market for Molson Coors, classified 30 years of the economy into regimes, and prepared four million rows of healthcare claims data for a GenAI product.

Most recent
Business Analyst Intern @ Abacus Insights
Looking for
2027 Business Analyst · Data Analyst · Consulting
View my work Résumé LinkedIn GitHub

Turn 01 Selected work

Three projects that explain what I do.

Economics, data science, and technical implementation — each one is a full case study behind the Details button.

Five-year forecast of the U.S. alcohol market by category add screenshot → assets/molson-coors.png
01 Forecasting · Economics · ML

Forecasted the U.S. Alcohol Market

The Data Mine × Molson Coors

How do macroeconomic conditions change what America drinks?

Five-year forecasts across Beer, Wine, Spirits, and RTDs. I owned Beer and RTDs — LASSO with macroeconomic drivers, chosen after SARIMAX and XGBoost failed on 35 annual data points.

Key findingRTDs are recession-resistant.

PythonRLASSOForecasting
Regime timeline of the U.S. economy since 1996 add screenshot → assets/macro-regime.png
02 Economics × Machine Learning

Macro Regime Dashboard

Independent · flagship

What state is the economy actually in, right now?

Gaussian Mixture Models on 30 years of FRED data classify every month since 1996 into Expansion, Slowdown, Stagflation, or Recession — then map which S&P sectors win in each.

Key findingCorrectly flags 2008, 2020, and the 2022 stagflation — unsupervised.

PythonGMMFRED APIscikit-learn
GitHub
Undercut success-probability heatmap from F1 telemetry add screenshot → assets/f1-undercut.png
03 Telemetry · Modeling · Strategy

F1 Undercut Strategy Analysis

Independent

When does pitting early actually gain you a position?

Nine visualizations from real race telemetry — stint degradation, pit timelines, and an undercut success-probability heatmap.

The questionDecide when the expected gain from acting exceeds the cost of waiting.

PythonFastF1pandasmatplotlib
GitHub

Turn 02 Selected experience

Real companies, real data, real deadlines.

Most recentJun – Aug 2026

Business Analyst Intern

Abacus Insights — healthcare data infrastructure

  • Profiled a 4M-row, ~300-column medical claims dataset in Databricks to diagnose what was missing and why — then built the specification the team's ticket backlog was cut from
  • Translated claims-processing rules from subject-matter experts into a probabilistic model and wrote a Python/PySpark generator that populated six fields across 4,027,321 rows with zero validation errors — repeatable on monthly refreshes
  • Sourced, validated, and documented authoritative healthcare reference code sets for a centralized data library, catching source and format traps before they corrupted downstream loads
Aug 2025 – May 2026

Project Assistant — U.S. Alcohol Market Forecast

The Data Mine × Molson Coors

Owned the Beer and RTD categories in a five-year market forecast. Selected LASSO with macroeconomic drivers after benchmarking SARIMAX and XGBoost on 35 annual observations; presented at the Data Mine Showcase.

Aug 2024 – May 2025

Project Assistant — Customer Segmentation

The Data Mine × AgReliant Genetics

Reduced 80+ raw features to ~20 modeling features and benchmarked K-means, DBSCAN, and hierarchical clustering with silhouette validation. 88% accuracy on real agricultural sales data, delivered to executives in Tableau.

Jun – Jul 2025

Data Science & Software Engineering Intern

Cygnet.One — Ahmedabad, India

Migrated legacy ASPX modules to ASP.NET MVC and built KPI dashboards for HR, managers, and employees — cutting manual reporting time by 40%.

Leadership & experience
CurrentAug 2026 – Present

Board Member — Student Fee Activity Board

Purdue Student Government

Reviewing 15–30 funding applications a cycle and deliberating allocation of a multi-million-dollar student activity fee. Decisions on criteria, not preference.

CurrentAug 2026 – Present

Teaching Assistant — ECON 260 Statistical Inference

Purdue University

Office hours, review sessions, and grading across probability, hypothesis testing, and regression.

CurrentSep 2026 – Present

Tour Guide

University Residences Guest Center, Purdue

Lead campus tours for prospective students and families, covering University Residences housing and dining, and host at the Guest Center between tours.

CurrentMay 2025 – Present

Treasurer

F1@Purdue

Budgets, expense tracking, and grant allocation for a $5,000+ annual budget. The other side of the table from SFAB.

Aug 2025 – May 2026

Teaching Assistant — ECON 252 Macroeconomics

Purdue University

Supported 120+ students through office hours, grading, and supplemental materials.

Jan – May 2025

Social Media Lead

Purdue SEARCH Club

Digital engagement and outreach.

2023

Team Lead

TiE TYE Program — Ahmedabad

Led a startup team from ideation to pitch; final round among 10+ teams.

Jun – Jul 2023

Tutor

Shyam Sir Classes — Ahmedabad

Taught Java to a class of 30 with a 100% pass rate, the summer before Purdue.

Turn 03 The grid

Came for the code.
Stayed for the Fed.

I came to Purdue from Ahmedabad in 2023 as a CS major, convinced I'd spend my career building software. Then I took macroeconomics.

Watching a single Fed decision move markets made me realize I didn't just want to build systems — I wanted to understand the systems they operate in. So I added Economics.

Since then my work has lived at the intersection: using code and data to understand markets, businesses, and the decisions inside them.

What I work on

01

Data → Decisions

Turning messy datasets into analysis people can actually act on. Most of the work happens before the model.

02

Economics × ML

Statistical and machine-learning methods applied to markets and economic behavior — regimes, forecasts, segmentation.

03

Business × Technology

Translating ambiguous business asks into scoped technical work, and technical findings back into language a stakeholder can use.

04

Systems

The pipelines and infrastructure underneath the analysis — built to be re-run by the next person, not thrown away.

Turn 04 The garage

More from the workshop.

Smaller projects, experiments, and systems work.

P-04Risk

Monte Carlo Portfolio Simulation

What's the worst that could plausibly happen to this portfolio?

10,000+ simulated paths via Geometric Brownian Motion on real market data — SPY, QQQ, DIA, single names. VaR, maximum drawdown, and full probability distributions.

PythonNumPyyfinance
GitHub
P-05Telemetry

F1 Race Replay

Can you rebuild a whole Grand Prix from the data alone?

Reconstructs full races from telemetry into an interactive 25 fps replay — live leaderboards, per-driver speed, gear, and DRS. Multiprocessing cut compute time by 80%.

PythonFastF1ArcadeNumPy
GitHub
P-06Segmentation

Customer Segmentation — AgReliant

Which of these customers actually behave the same way?

80+ features down to ~20. Four clustering approaches benchmarked with silhouette validation. 88% accuracy on real agricultural sales data, told in Tableau.

PythonRSnowflakeTableau
P-07Systems

Custom Unix Shell

How does the thing you type commands into actually work?

Pipelines, I/O redirection, signal handling, wildcards, subshells, and an interactive line editor with history and tab completion.

C / C++Unix
P-08Enterprise

KRA-KPI Management Platform

How do you make goal-tracking something managers actually open?

Role-specific interfaces for goal setting, KPI tracking, and end-to-end reporting, built at Cygnet.One. C# controllers wired to SQL stored procedures.

ASP.NET MVCC#SQL Server

Turn 05 Pit wall

What I'm actually strongest at.

Core

PythonSQLRJavaC / C++

Data & ML

pandasscikit-learnPySparkDatabricksSnowflakeTableau

Economics & Analytics

EconometricsForecastingRegressionFRED APIStata

Business

Business analysisData visualizationFinancial modelingAgile
+ full setup sheet

Languages

C#JavaScriptHTML/CSS

ML & Data Science

NumPyLASSOGMM / clusteringPCAXGBoostSARIMAX / time seriesfeature engineeringmatplotlib · seaborn

Big Data & Cloud

Spark SQLDelta Lake / MERGEUnity Catalogmedallion architecture

Finance & Markets

Monte Carlo / GBMVaR & riskregime classificationyfinancepanel dataSPSS

Healthcare Data

claims lifecycleadjudicationpayment integritycode sets (ICD · CPT · HIPPS · NDC)CMS geographic & rate files

Tools & Systems

Power BIExcelSQL ServerGit / GitHubJiraConfluenceUnix / LinuxASP.NET MVC

Turn 06 Podium

Achievements

2025 IMC Prosperity 3 Global Trading Challenge — Top 200 of 600+ teams

15-day algorithmic trading simulation with 12,000+ participants worldwide. Python trading algorithms across multiple asset classes, adapting strategy to competitive market dynamics in real time. The one that pointed me at markets for good.

2025 Federal Reserve Challenge — Purdue Team

National monetary-policy competition. Forecasts and policy recommendations, with research focused on distributional effects across socioeconomic groups, housing, and labor force participation.

2025 Moonshot Pitch Competition — Finalist

Final round among 20+ teams at the Purdue Innovates Incubator. Business case built on financial modeling, market analysis, and competitive positioning.

2024 Economic Scholars Program

One of 12–15 students selected. Payment systems, blockchain, and DeFi; designed a novel currency model and presented to faculty and industry professionals.

2024 Dean's List & Semester Honors — Spring 2024, Fall 2024

Purdue University.

+ more
2024 StarkHacks — Hardware Procurement

Coordinated hardware procurement and built the cost model behind it, with Indiana tax calculated in.

2023 Chhatra Sansad MUN — Verbal Recognition for Debate

Ahmedabad. Arguing a position under pressure — which turns out to be most of consulting.

Turn 07 Off the clock

  • teamMcLaren — papaya since before it was winning
  • sportsCricket · hurling
  • gamesValorant
  • hobbyOptimizing credit-card points like it's a portfolio
  • basedWest Lafayette, IN · from Ahmedabad

Turn 08 Final corner

Box, box.

Have a problem worth modeling? I'm recruiting for 2027 Business Analyst, Data Analyst, and consulting roles at the intersection of technology, analytics, economics, and strategy.

jhaverin@purdue.edu Résumé LinkedIn GitHub