SELECT name, role, years, focus FROM analysts WHERE metrics_are_trusted = TRUE AND dashboards_get_used = TRUE AND city = 'Philadelphia';
| name | role | years | focus |
|---|---|---|---|
| Nakshatra Agarwal | Senior Data Analyst | 5+ | BI, AML, forecasting |
Every number here comes from work I owned end to end. The case studies below explain how a few of them happened.
| Result | What changed | Where |
|---|---|---|
| $1.2M | Budget reallocated based on growth insights from healthcare product analytics | E-Solutions, 2025 |
| +18% | Conversion lift after I built the KPI framework and A/B testing setup from scratch | E-Solutions, 2025 |
| $175K | Excess cost cut per year by forecasting models that improved accuracy 15% | E-Solutions, 2024 |
| −27% | Churn, after funnel and cohort analysis found what was driving it | Spectrum WebApps, 2021 |
| 2 | Suspicious activity reports supported by anomaly analytics I built on Snowflake and AWS | Protiviti, 2025–26 |
| −35% | Documentation time, after bringing AI tools into SQL and documentation work | Protiviti, 2026 |
Each one follows the same path: the question someone needed answered, the data I had, what I did with it, and the decision it led to.
Consulting, product and compliance analytics, in that order.
AML compliance analytics on Snowflake and AWS, dbt models for AML features, and regulator-facing Tableau dashboards. Moved Excel reporting to automated SQL workflows.
Healthcare product analytics, KPI frameworks and A/B testing. Owned the monthly reporting cycle through to VP-level presentations.
Forecasting models and self-serve Tableau and Power BI dashboards for Finance, Marketing and Product, which shortened executive decision cycles by 32%.
Funnel and cohort analysis that cut churn 27%, and reporting frameworks that made product teams 30% more self-sufficient.
SQL and BI dashboards for healthcare and life sciences clients across 15+ accounts, plus VBA automation and time-series forecasts.
Built to get hands-on with tools before I needed them at work.
End-to-end analytics engineering: distributed cleaning in PySpark, warehouse models in Snowflake and dbt, automated quality checks.
ARIMA and ETS compared across 5 stores, with hypothesis tests to measure what promotions actually do to sales.
Event-driven pipeline that cleans inventory data and sends low-stock alerts. Tested locally with moto at zero AWS cost.
Earlier research: cell segmentation with SVMs (98.3% training accuracy) and sentiment analysis of film reviews weighted by each character’s impact.
Grouped by the part of the job they cover.
I’m open to senior data analyst, BI and analytics engineering roles. If your teams report five different numbers for the same thing, I’d like to hear about it.