Context
Enterprise reporting explains what happened. This work explored a future-state market-access platform that could help users reason about what might happen next and how uncertain those forecasts were.
My role
Product strategy, discovery, and a predictive-intelligence proof of concept for an anonymised healthcare client. I translated analytical methods into a future-state data-product vision.
Approach
Explored ARIMA and ETS forecasting, seasonality, and a triple-exponential-smoothing Excel demonstration. Six months was the preferred horizon in the source material, with 95% confidence-interval bounds proposed to communicate uncertainty.
Results
The framework considered monthly actuals refresh, model-error calculation, and adjustment. Exploratory regression considered relationships among market volumes, patients, claims, and initial out-of-pocket costs. The future-state work received recognition.
Limitations
This was exploratory product and analytical work, not a claim of production forecast accuracy or clinical impact. Longer horizons carry greater uncertainty. Client identities, private URLs, patient data, and original presentation material remain private.