# Model Studio
> CE and budget-impact models built once, defended everywhere.
Source: https://www.visfo.health/product/model-studio
Tags: Platform, AI-assisted

_CE and budget-impact models built once, defended everywhere._

A hosted modelling environment for cost-effectiveness and budget-impact models, transparent assumptions, country-ready inputs, version history, reviewer-friendly outputs and live wiring to the NMA Engine, GVD Workbench and HTA Console.

## What it is

The single hosted environment for the cost-effectiveness and budget-impact models that defend the price. One canonical CE model and one canonical BIM per indication, country-localised through controlled forks, with every assumption tagged to its evidence source and every reviewer-facing output generated from the live model rather than a static PDF.

It is not a forest of Excel files emailed between consultants and reviewers. Every input reads from a canonical source, NMA effect from the NMA Engine, cost from the country cost layer, utility from the utility library, every assumption carries its evidence reference and its sensitivity range, every run is versioned and reproducible, and reviewer-facing outputs are generated on demand rather than copy-pasted into a dossier.

## How we run it

1. **Stand up the canonical model** — Build the CE and BIM core once with assumptions tagged to evidence, inputs wired to canonical sources and a sensitivity frame the reviewer will recognise. No untraced numbers, no hidden cells.

2. **Localise through forks** — Country teams fork the core into a controlled localisation with country costs, utilities, comparator anchor and time horizon. The fork keeps the lineage to the core, so a core update propagates with diffs surfaced.

3. **Run sensitivity that survives review** — One-way, multi-way and PSA runs generated from the live model. Tornado diagrams, CEAC and scatter outputs render on demand with the seed and the assumption set recorded against each run.

4. **Ship reviewer-ready outputs** — Reviewer packs (model report, technical appendix, reproducibility seed) generated from the live model. Outputs change when the model changes; the dossier never lies behind the assumption set.

## What you get

- Canonical CE and BIM models per indication with assumptions tagged to evidence sources

- Country-localised forks with lineage to the core and diff surfaced on every core update

- Reproducible sensitivity runs (one-way, multi-way, PSA) with seeds and assumption sets recorded

- Reviewer-ready model report, technical appendix and reproducibility pack generated from the live model

- Live wiring into the Evidence Synthesis Engine, NMA Engine, GVD Workbench, HTA Console and Reimbursement Strategy Console so the model stays the canonical source of the deal math