hey there!

I'm Ritwik. I make data make sense.

I'm an analytics consultant in Princeton, NJ. By day I build forecasts, ML alerts and dashboards for Fortune 500 pharma teams at PharmaACE, after a couple of years doing the same for finance and supply chain clients at PwC. By weekend I'm usually on a trail somewhere.

See what I've built How I build with AI LinkedIn ↗ GitHub ↗

trail day. hat non-negotiable.

$300M+in revenue decisions shaped by my forecasts

40%+sharper alert timing from an ML smart-alert system

100+dashboards built (I stopped counting after that)

3+ yrsacross Big Four consulting and pharma analytics

the short version

I started out in a biotech lab writing Python to read DNA sequences. Turns out I liked the data more than the pipettes.

That led me to PwC, where I built dashboards and forecasting models for Fortune 500 companies, then to a Master's in Business Analytics at Maryland, and now to PharmaACE, where I help pharma brand and sales teams decide where to spend their time.

The common thread: I like sitting between the data and the people making the call, and translating in both directions.

the trail so far

Where I've worked

Pick what you care about and I'll highlight the relevant bits.

show me:

PharmaACE

Jun 2025 – present · Princeton, NJ

Consultant, Advanced Analytics

where I learned what a "line of therapy" is

Helped design an ML-driven smart-alert system for a cell therapy brand, swapping yes/no trigger flags for predictive models on patient journeys. Alerts now land 40%+ better timed, on a $250M portfolio.
Built forecasting models in Python for two flagship products, improving accuracy by 6% and feeding $300M+ in revenue and supply decisions.
Go-to analytics partner for Incentive Compensation, Sales Ops and Brand teams at a Fortune 500 pharma client, with 30+ recurring and ad-hoc analyses on claims, sales and distribution data.
Run weekly KPI reporting for 4 priority indications and lead a 3-person offshore team. Reporting SLAs are up 20%.
Automated SQL and Python pipelines over multi-million-row datasets, so reports go out 10% faster.

University of Maryland

2024 – 2025 · College Park, MD

MS, Business Analytics · Terrapin Scholar

went back to school to level up on ML

Machine learning, big data, decision science. Built SmartFlow, TransitIQ and a few other projects you'll find below.

PwC

Jul 2022 – May 2024 · Mumbai

Associate, Advisory · Advanced Analytics & Cloud

learned to explain a model to a CFO in 5 minutes

Led data architecture and dashboards across Finance, Sales, Inventory, Supply Chain and HR for Fortune 500 clients, cutting decision time by 200%.
Built a first-of-its-kind What-If dashboard in SAP Analytics Cloud, so execs could tweak assumptions and see the impact live.
Turned 20 years of financial history into a forecasting model and Finance cockpit for India's largest tractor maker. Reporting time dropped by half.
Built a Random Forest churn model for a Fortune 500 robotics company that lowered customer acquisition cost by 8%.
Automated ETL on SAP BW/4HANA across 100GB+ of client data: 60% faster processing, 35% fewer report errors.
Led 5 associates to build a 100+ dashboard demo library that helped win $500K+ in new work.
Modeled benefits cost and attrition for Health Industries clients and supported actuarial work on pricing and risk adjustment.

NIT Karnataka

2020 – 2022 · Surathkal, India

M.Tech Biotechnology · Bioinformatics intern

the "pipettes" era

Wrote BioPython pipelines to parse FASTQ/FASTA files for sequence alignment and mutation prediction on CRISPR data, speeding up gene detection by 20%.

SRM Institute

2016 – 2020 · Chennai, India

B.Tech · First Class with Distinction

the trailhead: where it all started

things I'm proud of

A few problems I've cracked

Client names stay private. The results are real.

pharma · cell therapy

40%+better-timed alerts

Alerts that show up before you need them

Predictive models on patient journeys tell reps which doctor to call and when, instead of after the fact.

PythonStats modelingClaims data

pharma · brand strategy

$300M+in decisions informed

Forecasts people actually plan around

Time-series and regression models for two flagship drugs, used to set sales and supply plans across regions.

Time seriesRegressionSQL

consulting · Fortune 500

1stof its kind at PwC

The What-If dashboard

Execs drag a slider on hiring or inventory and watch the numbers move in real time.

SAP Analytics CloudPipelines

manufacturing · finance

50%less reporting time

20 years of data, one cockpit

A forecasting model plus pro-forma views that gave leadership a look ahead at cost and revenue.

PythonSACForecasting

robotics · customers

8%lower acquisition cost

Catching churn early

A Random Forest model that showed sales which accounts to save and where to stop spending.

scikit-learnRandom Forest

PwC · internal

100+dashboards in the library

The demo library

Led a team of five to build a cross-industry dashboard library that cut turnaround by 70% and helped win new clients.

Power BITableauQlik

my secret weapon

Building with AI

I'm not an AI researcher. I'm the person who gets AI to ship real things. At PharmaACE I've led several initiatives where Claude and AI agents do the heavy lifting and I design the workflow, check the output and get it in front of users.

Auto-refreshing newsletter

AI agents pull the latest industry news, fact-check each item, write the newsletter and email it out on a schedule, with no manual copy-paste.

Fetch news→Fact-check→Write→Email

See it live: MathCo Rx Weekly

Whitepaper generator

Turns a topic and a pile of sources into a structured, on-brand first draft that the team refines instead of starting from a blank page.

Sources→Outline→Draft→Review

AI-powered EDA tool

Drop in a dataset and get the profiling, charts and first-pass insights in minutes, so analysts spend their time on the questions instead of the setup.

Upload data→Profile→Insights

Websites designed with Claude

Took sites from idea to a polished, mobile-ready build with Claude, including this one.

Idea→Design→Live site

Internal apps

Built working web apps for the team with Claude Code and Google AI Studio, going from a request to a usable tool far faster than a traditional build.

ReactClaude CodeAI Studio

Multi-agent workflows

Chaining AI agents so each one does a single job (research, write, verify, deliver), with human checkpoints where accuracy matters.

AgentsHuman-in-the-loop

StackClaudeClaude CodeAI agentsGoogle AI StudioReactPrompt designAutomated fact-checking

the trick isn't knowing how AI works inside. it's knowing what to ask it, and checking what it gives back.

side quests

Projects

Grad school work and things I built for fun. Code's on GitHub.

MathCo Rx Weekly passkey

An AI-run pharma newsletter, built as a concept for MathCo's pharma wing. Claude finds the week's news and writes the briefs; an editor ships it in one click.

ReactClaude APINetlify

locked: ask me for the passkeyNot an official MathCo project

SmartFlow ↗

Linear programming that picks the cheapest mix of road, rail and air to move goods between cities, without breaking time or storage limits.

PythonLinear programming

TransitIQ for Amtrak ↗

Where Amtrak could earn more, carry more riders and run leaner, backed by SQL analysis and Tableau dashboards.

SQLTableauPython

Rank-based recommender ↗

Simulates shoppers and recommends products ranked by popularity and personal taste.

PythonPandasML

Pharma KPI & forecast sim ↗

A mock pharma commercial database with KPI queries and ARIMA / Holt-Winters demand forecasts.

SQLARIMATableau

what's in my backpack

Toolkit

data science

Predictive modeling
Time-series forecasting
Regression
Random Forest
Hypothesis testing
Linear programming

code & data

Python
Pandas · NumPy
scikit-learn
SQL
R
SAS
Databricks

dashboards & cloud

Tableau
Power BI
Qlik Sense
SAP Analytics Cloud
SAP BW/4HANA
AWS
Azure

domain

Incentive comp
Claims data
Pharma commercial
Finance
Supply chain
Stakeholder wrangling

when I'm not modeling data

Off the clock

Hiking & national parks

Big on long trails and bigger views. Always planning the next park.

bucket list: 46 places and counting

Running

Building my aerobic base one long run at a time. Yes, I track my VO2 max. It's data.

longest run so far: ~15 km

Racquets & cricket

Tennis, badminton and pickleball on weekends. Cricket whenever there's a match on.

yes, I'll analyze your batting average

Board games

Catan, Wingspan, Everdell and Azul. Strategy games are just optimization problems with snacks.

will trade sheep for wood

Building with AI

Tinkering with Claude Code and Google AI Studio on side projects and agent ideas.

this site included

Fun fact

I have a published paper on turning slaughterhouse waste into biodiesel. Ask me about it.

Springer, 2020

schooling

Education & honors

2025

MS, Business Analytics

University of Maryland, Robert H. Smith School of Business

Terrapin Scholar

2022

M.Tech, Biotechnology

National Institute of Technology Karnataka, Surathkal

GATE Scholar · thesis on CRISPR-directed transposons

2020

B.Tech

SRM Institute of Science & Technology

First Class with Distinction

GATE 2020: 96th percentile, All India Rank 538 Rajya Puraskar: 8 years of community service with the Scouts & Guides Published: BCAB, Springer (2020)

let's talk!

Got a data problem? Or a good trail?

I'm open to analytics, data science and consulting roles. I'm happy to chat over coffee, virtual or otherwise.

LinkedIn in/ritwikp ↗ GitHub RitwikPandeyy ↗