Product, strategy & business operations. Computer Science at UW–Madison, business master's at Duke Fuqua. I turn ambiguous problems into working processes and data-backed decisions.
A model is only worth something if someone can act on it. I care about the part after the analysis — the decision it unlocks, the process that still works once I hand it over, and the number that actually moves.
A computer science degree, a business master's, and a company I built before either of them.
Business and product operations professional with a Computer Science degree from UW–Madison and a business master's from Duke Fuqua — skilled in SQL, Python, Excel, and Tableau to turn ambiguous problems into working processes and data-backed decisions.
At 17, I founded and scaled an education product to 4,000+ customers with no funding and no team. That's still the most useful thing on my résumé: it's where I learned to size an opportunity, build the thing, sell it, and keep it running at the same time.
I'm currently a business analyst at Storied Data, turning stakeholder requirements into dashboards and reporting on their BI platform. Before that I owned KPI tracking, market analysis, and go-to-market work across two internships and a graduate consulting practicum — building financial models and business cases, then translating them into recommendations executives can actually act on.
Founded and scaled an education product to 4,000+ customers with no funding, no team, and no safety net. I built it, sold it, supported it, and learned how a business actually holds together — the reason I do operations work the way I do now.
A cohort and churn analysis on an open SaaS dataset, answering the question every growth team argues about: which acquisition channel brings customers who actually stay. Ends in one recommendation, not a wall of charts.
An interactive model that sizes the business case for automating a support queue — ticket volume, deflection rate, cost per ticket, payback period. The framework I built at Troon, generalised so anyone can run their own numbers.
A full CAC, LTV, and payback breakdown of a real consumer brand using public filings and clearly-labelled assumptions, ending in a verdict on whether the model works at scale.
A reusable template for turning a vague stakeholder request into a dataset spec, reporting logic, and a dashboard wireframe — the translation layer that decides whether a BI project ships or stalls.
One analysis presented three ways: a one-page memo, a five-slide board readout, and a two-minute verbal version. Built to show the skill nobody puts in a portfolio — making a number land with the person who has to act on it.
TAM, SAM, and SOM built bottom-up rather than borrowed from a press release, plus a competitive landscape and a go/no-go recommendation with the conditions that would change the answer.
A Tableau build over a messy operational dataset — reconciled, validated, and modelled first, so the dashboard answers questions instead of raising them. The data cleanup is the actual project.
A small web tool that turns a north-star metric into a driver tree — every input, who owns it, and how far it would need to move to matter. Built because most KPI decks list metrics without showing how they connect.
A scenario model showing what happens to revenue, mix, and churn when you move a tier boundary — the decision people usually make on instinct because the spreadsheet was too painful to build.
A model over usage and support signals that flags accounts likely to leave, wrapped in a view a CSM could actually use on a Monday morning. Interpretability over accuracy — a score nobody trusts changes nothing.
A repeatable framework for finding which manual processes are worth automating, with a time-saved calculator attached. Formalising the approach that cut manual work 40% running a 100+ member organisation.
The overlap between someone who can write the query and someone who can defend the recommendation.
Financial models, business cases, opportunity sizing, and scenario analysis — built to survive a room full of executives asking why. At Duke's client consulting practicum I built a KPI framework and cost model that cut average resolution time by more than a day.
A CS degree means I can pull the data myself, read the system that produced it, and talk to engineers without a translator in the room.
Process maps, standardized reporting, and automation — I cut manual work by 40% running operations for a 100+ member organization.
Resolution time, deflection rate, cost per ticket, retention. If it isn't measured, it isn't managed — so I define the metric before the initiative.
Hospitality, financial services, IT consulting, and now a BI platform — four industries in four years. I've had a lot of practice getting useful before I'm comfortable.
I've been the liaison across 10+ stakeholders in sales and operations, cutting clarification cycles by 30% simply by making sure everyone was working from the same document.
Every skill below has a project attached to it — see Work.
Four stages. Each one ends with something a decision-maker can look at.
Talk to the people closest to the problem, map the process as it actually runs, and write down the question we're really answering. Most bad analysis is a good answer to the wrong question.
Pull it myself where I can — SQL against the operational data, market and competitor research where internal data runs out. Assumptions get labelled as assumptions, not smuggled in as facts.
A model with the levers exposed, scenarios run at the edges, and an ROI story that holds up when someone pushes on it. The output is a recommendation with a number attached, not a deck of observations.
Executive-ready reporting, a standard format the team can keep running after I leave, and KPIs that tell us within weeks whether the decision was right. Analysis that doesn't change behaviour was a hobby.
Business intelligence, strategy, and analyst work across consulting, operations, and go-to-market.
Two degrees on both sides of the technical/business line, plus the roles where I learned to run things.
Actively looking for full-time roles in product, strategy, and business operations — and happy to relocate anywhere in the US. Email is the fastest way to reach me.
arnav.bharadia@duke.edu