MedTWIN.AI
Available now

Research AI (DRE)

Transform messy clinical spreadsheets into audit-ready publications with transparent AI assistance.

DRE Research interface
Overview

From spreadsheet to publication

Research AI is designed for real-world clinical datasets: messy headers, multi-sheet workbooks, and evolving study definitions, while keeping decisions reproducible.

Explicit mapping & analysis specs
Deterministic statistics (no AI math)
Exportable artifacts + drafts
Audit trail for major actions
Research workflow

Complete research toolkit

Everything you need for an auditable workflow from input to export.

AI-assisted mapping

Raw → canonical transformation with explicit, reviewable specs.

Deterministic analysis

A stats engine you can rerun later and get the same results.

Draft generation

Structured narrative grounded in computed outputs, not fabricated math.

Audit trail

Traceable inputs, transforms, approvals, and exports.

Collaboration

Workflows designed for teams and reviewers.

See it in action

From data upload to publication-ready manuscripts.

Research workflow pipeline

Complete workflow from mapping to publication

ML analysis with feature importance

ML baselines with interpretable feature analysis

Cohort characteristics table

Publication-ready cohort tables

AI-generated paper introduction

AI-drafted manuscripts grounded in your data

How it works

01

Upload data

Import spreadsheets and cohort exports

02

Map variables

AI proposes canonical mappings; you approve

03

Run analysis

Deterministic engine computes results

04

Export

Artifacts + drafts with provenance

Supported study types

  • Retrospective clinical cohort studies
  • Simple parallel-group RCTs
  • Case-control studies
  • Cross-sectional studies

Technical specs

InputExcel, CSV
OutputArtifacts + drafts
APIREST (docs)
GovernanceAudit trail + approvals

Research AI is built to reduce "black box" risk: deterministic computation, explicit specs, and traceable outputs.

Ready to try Research AI?

Launch the app, or request a demo for your dataset and study type.