What we offer | ClinAI

Research Workspace

A structured environment for your healthcare AI project

ClinAI provides a dedicated, secure research workspace that brings the key components of a healthcare AI project together in one place.

Each approved project has its own structured workspace, connecting project information, research data, computing resources, model development, validation and research outputs.

Within your workspace, you can:

Organise your research through defined project stages, from setup to completion.
Maintain project, dataset and governance records.
Access authorised research data and computing resources through AWS.
Organise research code, datasets, models and outputs in designated project folders.
Track research progress, activities and documentation.
Manage access for approved research collaborators.
Maintain a structured record of your project as it develops.

Designed for healthcare AI research

ClinAI combines a web-based research portal with an AWS research environment.

The portal supports project organisation, documentation and progress tracking, while the AWS environment provides access to research data, computational resources and tools for analysis and AI model development.

Researchers receive a step-by-step onboarding guide when they begin their approved project, including instructions for accessing their workspace, running analyses and saving research outputs.

ClinAI provides the research environment, infrastructure and project structure. Researchers retain responsibility for their research questions, methods, scientific decisions, data management, validation and interpretation of results.

Guided Research Workspace

Guidance throughout your healthcare AI research journey

ClinAI provides structured guidance to help researchers navigate the practical and technical aspects of healthcare AI research, from project setup through analysis, validation and reporting.

The Guided Research Workspace combines a structured project workflow, step-by-step onboarding instructions and technical guidance to help researchers work within their approved research environment.

How ClinAI Supports Your Research

Understand — Review your research objectives, study protocol, published methods or existing code to identify the requirements of your project.
Plan — Organise your research into defined stages, activities and expected outputs.
Prepare — Identify the required datasets, software, computing resources and project workspace configuration.
Execute & Troubleshoot — Follow practical guidance for accessing your research environment, running scripts, managing files and resolving technical issues.
Review & Interpret — Examine intermediate results, identify potential errors and assess whether the workflow is progressing as intended.
Document — Maintain structured records of research methods, datasets, software environments, code modifications, validation activities and outputs.
Reproduce — Support reproducibility by retaining the information needed to review, repeat and validate research workflows.

Step-by-Step Researcher Onboarding

Researchers receive a practical onboarding guide when their project workspace becomes available.

The guide explains how to:

Access the ClinAI research environment through AWS.
Locate authorised research datasets and project folders.
Connect to computing resources and open their project workspace.
Run research scripts and manage analysis outputs.
Save research files and verify storage in Amazon S3.
End computing sessions safely.

The guide is designed for researchers who may be unfamiliar with AWS, while allowing them to use their existing research and analytical expertise.

Researcher-led, ClinAI-guided

ClinAI provides the research environment, structured workflow and technical guidance. Researchers remain responsible for their research questions, scientific methods, analytical decisions, validation and interpretation of results.

ClinAI supports the research process without replacing researcher oversight or independent scientific judgement.

Compute Included

Computing resources to support your research

ClinAI provides access to cloud-based computing and storage within an approved research workspace, allowing researchers to undertake healthcare AI projects without having to independently establish and manage AWS infrastructure.

Computing resources are configured according to project requirements and the agreed workspace allocation.

Your workspace can include:

Standard CPU Compute — Computing resources for data preparation, preprocessing, statistical analysis, routine machine-learning workflows and model evaluation.
Project Storage — Authorised Amazon S3 storage and designated working directories for research datasets, scripts, models, intermediate files and outputs.
Configured Research Environment — Access to Python, AWS utilities and other approved software tools required for the research project.
GPU Compute — Accelerated computing for projects requiring more intensive model training or evaluation. GPU resources are arranged separately according to project requirements and applicable charges.
Scalable Resources — Additional computing capacity, storage or specialised software environments can be arranged when research requirements exceed the standard allocation.

Focus on the research, not the infrastructure

ClinAI provides and administers the underlying AWS research infrastructure, including project workspaces, access permissions and computing resources. Researchers can focus on their datasets, analytical methods, model development and research outcomes rather than independently managing cloud infrastructure.

Project-Specific Resource Allocation

Computing and storage requirements vary between research projects.

During project onboarding, ClinAI works with researchers to identify suitable resources, software requirements and any additional computing needs.

Standard resource allocations and any additional charges are agreed upon as part of the project setup.

Model Development

Develop and evaluate models within your research workspace

ClinAI provides a structured computing environment for developing, testing and evaluating healthcare AI models while keeping research activities connected to their datasets, methods and project documentation.

Researchers can develop models using their own code, algorithms and analytical approaches within their authorised project workspace.

Within your workspace, you can:

Develop Models — Build, train and test machine-learning and AI models appropriate to your research question.
Compare Approaches — Evaluate different algorithms, model configurations and experiments within the same research project.
Manage Model Versions — Organise model iterations, configuration files and associated research outputs to support version tracking.
Connect Models to Data — Maintain records linking datasets, preprocessing methods, model configurations and resulting outputs.
Record Performance — Save model evaluation metrics, predictions, tables, figures and experimental results.
Prepare for Validation — Progress selected models into a structured validation workflow, including assessment of performance, limitations and reproducibility.

From experimentation to a documented model

Healthcare AI model development often involves multiple experiments, adjustments and iterations.

ClinAI provides designated project folders for research code, working datasets, models, logs and outputs, together with a structured documentation workflow.

This helps researchers organise their experiments and maintain the information needed to understand which model was developed, using which data and methods, and how it performed.

Where enabled, ClinAI research documentation automation can support the recording of model development activities and the preparation of structured project records.

Researcher-Led Model Development

Researchers remain responsible for selecting appropriate methods, developing and testing their models, interpreting performance results and determining whether models are suitable for their intended research purpose.

ClinAI provides the research infrastructure, project organisation and documentation support rather than replacing scientific judgement or independent model validation.

Reproducibility Tools

Make your research easier to understand, review and reproduce

Healthcare AI research involves more than producing a final model or analysis result. Datasets, preprocessing methods, code, software environments, model configurations and research decisions can all influence the findings.

ClinAI provides a structured research environment and documentation tools to help researchers maintain a traceable record of their work throughout the project lifecycle.

ClinAI Supports Reproducibility Through:

Code and Script Records — Organise and retain the scripts, notebooks and code used at different stages of the research project.
Software Environment Documentation — Record the software, packages, dependencies and versions required to execute research workflows.
Data Processing Records — Document dataset preparation, cleaning, transformations and preprocessing methods.
Model Version Records — Maintain information connecting model iterations with the datasets, code, configurations and outputs used during development.
Research Modifications — Record changes made to analytical methods, code or model configurations, including deviations from published methods when reproducing existing studies.
Research Provenance — Maintain links between research data, analytical methods, model development, validation and resulting outputs.
Workflow Reproduction — Preserve the information needed to review, repeat or independently reproduce research analyses.

Reproducibility built into the workflow

Rather than reconstructing the research process after a project is completed, ClinAI encourages researchers to maintain reproducibility records as their work progresses.

Research files are organised within designated project folders, while structured documentation supports the recording of methods, technical configurations, analytical decisions and outputs.

Where enabled, ClinAI research recording automation can assist with capturing selected execution activities, code changes, model development records and outputs.

Researchers remain responsible for documenting scientific decisions, reviewing records and ensuring that sufficient information is retained to reproduce their work.

Supporting Transparent Healthcare AI Research

By connecting research activities with structured documentation, ClinAI helps researchers prepare their work for internal review, independent validation, scientific publication and future research.

Automated Research Documentation

Build your research documentation as the project progresses

ClinAI provides a structured documentation workflow that helps researchers organise and maintain their research records throughout the project lifecycle.

Rather than reconstructing methods, technical configurations, research activities and decisions at the end of a study, ClinAI supports the progressive development of project documentation as research is undertaken.

ClinAI Supports Research Documentation Through:

Structured Project Records — Organise documentation covering project setup, research data, analysis, model development, processing, validation, outputs and governance.
Reusable Project Information — Use existing project details and research records across relevant documentation sections, reducing repeated data entry.
Research Activity Recording — Where enabled, incorporate selected execution activities, technical records, model development information and research outputs into project documentation.
Changes and Decisions — Maintain records of important modifications, methodological decisions and their rationale.
Draft Documentation — Generate structured documentation drafts from available project information for researcher review and completion.
Documentation Review — Allow researchers to review, edit and approve documentation before it is treated as a final research record.
Research Reporting — Maintain organised records that can support methods reporting, validation, research governance and scientific publications.

A Structured Research Documentation Workflow

ClinAI organises research documentation into seven stages:

Project Setup — Research objectives, investigators, project details and governance information.
Research Data — Dataset descriptions, sources, permissions and cohort information.
Analysis and Models — Analytical methods, algorithms, model configurations and evaluation plans.
Research Processing — Execution activities, processing steps, technical records and intermediate outputs.
Validation — Data quality checks, model evaluation, reproducibility assessments and limitations.
Research Outputs — Results, reports, model artefacts and final research deliverables.
Governance and Closure — Research oversight, documentation review, data management and project completion.

Documentation That Develops With Your Research

ClinAI combines information entered by researchers with available technical records and automatically generated documentation.

Where research recording automation is enabled, selected activities and outputs can contribute to the project documentation workflow.

Researchers remain responsible for reviewing the accuracy and completeness of their records, documenting scientific decisions and approving final documentation.

Researcher Control

ClinAI supports documentation without replacing researcher oversight.

Automatically generated or AI-assisted drafts are intended for review and editing before becoming part of the approved project record.

Governance & Provenance

Keep data, governance and research decisions connected

Healthcare AI research requires clear documentation of where data originated, how it may be used, how it has been processed and what governance arrangements apply.

ClinAI provides a structured framework for maintaining this information throughout the research lifecycle, connecting data governance and provenance with project activities, model development, validation and research outputs.

ClinAI helps you:

Register Data Sources — Document dataset origins, custodianship, permitted uses and relevant data descriptions.
Document Permissions and Access — Maintain records of data access requirements, approved researchers and project-specific access arrangements.
Record Ethics and Governance — Document ethics approvals, consent arrangements or waivers, institutional authorisations and applicable research governance requirements.
Track Data Preparation — Record data cleaning, cohort selection, preprocessing, transformations and the creation of derived datasets.
Maintain Data Lineage — Document relationships between source datasets, processed data, analytical methods, model versions and resulting outputs.
Record Researcher Roles — Maintain information about project investigators, collaborators, responsibilities and authorised access.
Identify Outstanding Requirements — Record missing documentation, pending approvals or governance matters requiring researcher or institutional attention.

Traceability From Data to Research Outputs

ClinAI supports a structured research record connecting:

Data Source → Data Preparation → Analysis → Model Development → Validation → Research Outputs

This helps researchers understand which datasets were used, how they were transformed, which analytical methods were applied and how the resulting models and research outputs were produced.

Where enabled, research activity recording and documentation automation can contribute technical information to the project record.

Governance Throughout the Research Lifecycle

Governance and provenance are integrated into ClinAI's structured project documentation workflow, from initial project registration through research completion and closure.

Researchers can maintain relevant governance information alongside their research records rather than treating these requirements as separate administrative activities.

Researcher and Institutional Responsibility

Researchers and their institutions remain responsible for obtaining and maintaining appropriate ethics approvals, data access permissions, privacy safeguards and other required authorisations.

ClinAI provides the infrastructure and documentation framework to support research governance and provenance. It does not replace institutional oversight, independently certify compliance or grant permission to use research datasets.

Model Validation

Evaluate Model Performance, Generalisability and Limitations

ClinAI provides a structured research environment for evaluating healthcare AI models beyond their initial development.

Validation activities, datasets, model versions, performance metrics and research findings can be organised within the project workspace, helping researchers maintain a traceable record of how their models were evaluated.

ClinAI helps you:

Define the Validation Approach — Document the model's intended use, target population, validation datasets, outcomes and evaluation methods.
Evaluate Model Performance — Record relevant performance measures such as discrimination, calibration, sensitivity, specificity, predictive values and other metrics appropriate to the research question.
Compare Model Versions — Assess different model versions using defined evaluation methods and datasets.
Conduct Internal Validation — Evaluate model performance within the development study using appropriate methods such as cross-validation, bootstrapping or held-out test datasets.
Support External Validation — Evaluate models using independent datasets, patient populations, institutions or research collaborators, subject to appropriate data access and governance approvals.
Assess Subgroup Performance — Examine model performance across clinically relevant patient groups to identify potential differences in accuracy, calibration or generalisability.
Document Limitations — Record methodological limitations, missing data, potential bias, dataset differences and other factors affecting interpretation.
Maintain Validation Records — Connect validation methods and results to the specific model version, dataset, code and computational workflow evaluated

From Model Performance to Validation Evidence

A model performing well during development does not necessarily mean it will perform equally well in different patient populations, healthcare institutions or clinical settings.

ClinAI supports the progressive development of a structured validation record that documents:

Model Version → Intended Use → Validation Population → Dataset → Evaluation Methods → Performance → Subgroup Results → Limitations

This helps researchers understand what was evaluated, how performance was measured, which populations were represented and what conclusions can reasonably be drawn from the findings.

Supporting Transparent Healthcare AI Evaluation

ClinAI provides designated project folders and structured documentation for organising validation methods, results, figures, reports and related research outputs.

Where enabled, research documentation automation can assist with recording selected technical activities and preparing validation documentation for researcher review.

Researcher-Led Validation

Researchers remain responsible for selecting appropriate validation methods, assessing the quality and representativeness of validation datasets, interpreting findings and determining whether their models are suitable for their intended research purpose.

ClinAI supports model validation and documentation but does not independently certify model safety, clinical effectiveness or readiness for clinical deployment.