Commercial Forecasting at Scale: Tanmay Sharma’s Approach to Accuracy, Levers, and Trust

Tanmay is a pharmaceutical demand forecasting and commercial analytics leader whose work has focused on building forecasting systems that operate like enterprise infrastructure: scalable across large portfolios, governed for auditability, and engineered to translate uncertainty into actionable commercial levers. Over an eight-year tenure at Gilead Sciences, Tanmay led demand forecasting for HIV treatment and prevention in a market that is large, mature, and relentlessly competitive. In that setting, forecasting is not a reporting exercise. It is a core operating discipline that can directly influence supply reliability, access planning, commercial strategy, and executive confidence.

Why scale changes what “good” looks like

HIV is often considered a useful environment for forecasting because it combines a mature treatment category with ongoing competitive activity. The market has more than a decade of commercial history, a deep set of branded and generic molecules, and a steady cadence of launches and competitive events. The market is also large, roughly $22B annually. At that scale, a 1% point error implies about $220M of mis-estimation. Errors of that magnitude are not academic. They shape decisions on manufacturing and inventory posture, field and marketing investment, payer strategy, and the credibility of long-range plans.

Equally important, HIV is a chronic market where performance is often decided by patient dynamics and share redistribution rather than pure category expansion. The commercial question is not only, “How big is the market?” It is, “How do starts, switches, persistence, and discontinuations translate into share and demand over time, especially when competitors shift behavior in specific segments?”

What Tanmay implemented: a forecasting system built for portfolio reality

Tanmay approached forecasting as a system design problem. The goal was to produce decision-ready forecasts at portfolio scale with repeatable refresh cycles and disciplined governance.

1) A portfolio-scale demand forecasting engine

Tanmay built and led a demand forecasting model that ran across the HIV portfolio, spanning many molecules and more than a decade of market history, while accommodating multiple launches and competitive changes each year. The model was designed to be scalable and consistent across brands and time horizons, so that portfolio updates did not require rebuilding logic from scratch each cycle.

Performance was managed as an operational metric, not a retrospective surprise. On an annual basis, the forecasting system was designed to keep error rates very low. In a category of this size, forecasting accuracy can play an important role in helping teams manage planning volatility and align across functions.  It also shifted internal conversations away from debating whether the number was credible and toward debating which levers the organization should pull.

2) A separate patient in-outs model for executive levers

In chronic therapy markets, leadership decisions often hinge on understanding flow and mix, not only totals. Demand models can be highly accurate yet still fail to answer the questions executives ask most:

• What happens to overall share if naive starts move by a small amount?
• What happens when switch dynamics shift between specific segments?
• How does persistence amplify a small share swing over time?

To address those questions, Tanmay implemented a separate patient in-outs model as a companion to the demand forecast. Its purpose was not to replace the demand engine, but to make the market’s mechanics legible. The patient flow view translated market share swings in naive and switch segments into overall share outcomes, which was especially important in a chronic category where switching and persistence determine long-run volume.

This companion model gave commercial leaders a transparent way to connect strategy to outcome. Instead of treating share changes as ex post explanations, the organization could evaluate prospective moves and understand which segments and behaviors would drive the biggest impact.

3) Simulation-based communication of uncertainty

In competitive categories, leaders do not only need an expected value. They need a credible range. Tanmay incorporated simulations into leadership discussions to visualize the distribution of plausible outcomes under different assumptions, sensitivities, and competitive responses.

This approach improved the quality of decisions in two ways. First, it reframed uncertainty as something that could be bounded and managed rather than ignored. Second, it created a common language across functions for discussing risk. The approach aimed to support faster scenario evaluation, clearer tradeoffs, and greater alignment on organizational assumptions.

How the system earned trust

Trust in a forecast is built through governance and repeatability. Tanmay implemented disciplined assumption management so that key drivers were explicit, versioned, and reviewable. Refresh cycles were designed to be repeatable, with quality checks that reduced noise and prevented unintended drift. The focus was not only accuracy, but explainability: what moved, why it moved, and which levers mattered most.

At portfolio scale, this governance is a strategic advantage. It reduces the operational burden on forecasting teams and reduces friction for stakeholders. When leaders see a forecasting system behave consistently across cycles, confidence compounds, and the forecast becomes a foundation for planning rather than a topic of negotiation.

Conclusion

Commercial forecasting at scale is not a modeling contest. It is the disciplined implementation of a system that performs under competitive pressure, scales across portfolios, and earns executive trust through accuracy and explainability. Tanmay’s work in HIV treatment and prevention forecasting illustrates this approach in practice, pairing detailed forecasting in a large, mature market with patient flow modeling and simulations intended to help teams navigate uncertainty. As pharmaceutical organizations push toward faster planning cycles and more rigorous decision support, this systems-based approach increasingly defines best practice.

This article is for informational purposes only and does not substitute for professional medical advice. If you are seeking medical advice, diagnosis or treatment, please consult a medical professional or healthcare provider.