Presentation (CTI)

OBJECTIVES: This study evaluates a modular, scenario-driven forecasting solution that enables transparent, explainable, and rapid revenue forecasting for pre-launch pharmaceutical assets. It addresses inefficiencies in current forecasting approaches characterised by fragmented assumptions, limited standardisation, and manual, expert-driven processes.
METHODS: A minimum viable product (MVP) was designed, leveraging multiple AI agents, ML algorithms and data pipelines within a scalable analytics platform, to support end-to-end forecasting workflows. Key capabilities include multi-source data extraction and integration, claims-based market sizing, analogue identification, event detection and quantification, uptake curve modelling, and scenario comparison. Forecasts are generated by integrating patient-level inputs, historical analogues, pricing data, and market events into structured models. The system supports generation and comparison of custom scenarios, with interactive refinement of assumptions and real-time recalculation of outputs.
RESULTS: The proposed solution enables rapid generation and comparison of revenue forecasts across multiple scenarios, improving transparency of assumptions and methodological traceability. By standardising inputs such as analogue selection and uptake modelling, the system reduces reliance on manual processes and improves consistency across analyses. Initial evaluation indicates reductions in forecast cycle time, enhanced ability to iterate scenarios, and improved user accessibility through guided workflows and integrated dashboards.
CONCLUSIONS: This modular forecasting approach demonstrates the potential to transform pre-launch pharmaceutical forecasting by enabling faster, more transparent, and scenario-driven decision-making. By integrating structured data inputs with user-driven scenario planning, the solution supports more robust strategic evaluation of pipeline assets. Future development will focus on expanding modelling sophistication (e.g., advanced simulation and sensitivity analysis) and scaling adoption across broader therapeutic areas and use cases.