START MIDD
Choose the programme and clinical asset
MIDD stays programme-native, but you should not need to know internal IDs. Search the governed portfolio, choose an asset, and continue where the team left off.
RECENT WORK
Continue a recent programme
Loading recent work…
MODEL-INFORMED DEVELOPMENT WORKSPACE
MIDD Command Center
Turn clinical-development questions into governed model evidence, inspectable uncertainty and accountable human decisions—without losing programme context.
Loading canonical programme context…
Question-led modellingTyped evidence provenanceIndependent human review
Programme statusPhase —
STEP 1 OF 5
Create the data spine
Use only human-QC-approved Clinical Development imports to create an immutable analysis snapshot.
Loading datasets…
STEP 2 OF 5
Declare the decision and analysis plan
Start with the clinical question and context of use. Engine selection follows the decision—not the reverse.
Loading plans…
Independent plan review
STEP 3 OF 5
Execute an approved plan
Bind the approved plan to an immutable dataset and exact engine version. A run produces COMPUTED—not REVIEWED—evidence.
Loading runs…
SCIENTIFIC RESULTS
Analysis workspace
Inspect stored COMPUTED output, diagnostics, uncertainty and exact provenance before human review.
Run an approved analysis or select a previous run to inspect its scientific result.
STEP 4 OF 5
Complete human model review
A successful run is not a credibility conclusion. Review must be independent of plan authoring and execution.
Loading reviews…
STEP 5 OF 5
Assemble the evidence package
Package qualified plans, data lineage, engine provenance, checks, contributions and limitations. This supports review; it does not imply regulator acceptance.
PACKAGE PREVIEW