Healthcare and Community Services
Give administrative time back to frontline services.
Administrative automation, policy search, workforce analytics and privacy uplift for healthcare and community providers, with clinical decisions left entirely to clinicians.
Operational context
How the sector actually runs.
Administrative load competes directly with service delivery. Every hour a coordinator spends assembling reporting evidence or chasing a credential renewal is an hour not spent on the people the service exists for.
Rostering is constrained by qualifications, not just availability. A shift cannot be filled by whoever is free; it needs someone with the right credential, the right clearance, the right training currency and often continuity with a specific participant or resident.
Accreditation and funding reporting demand evidence, continuously. Providers operating under aged care, NDIS or community service agreements must be able to produce policy versions, training records, incident documentation and compliance evidence on request, usually from several disconnected systems.
Policy documents are numerous, versioned and frequently updated. Staff need the current answer to a procedural question during a shift, and the intranet search that returns a superseded policy is worse than no search at all.
Privacy expectations are high and the data is sensitive by default. Even administrative records carry health information, and the governance obligations follow the data wherever it goes.
High-value problems
Where the money and the risk actually sit.
Our coordinators spend more time on reporting than on coordination
We automate evidence assembly from the systems that already hold it, so reporting becomes a review of a compiled pack rather than a manual collection exercise. A quality manager still signs off what is submitted.
Staff cannot find the current version of a policy when they need it
A governed policy search returns the current approved version with its effective date and a citation, and does not surface superseded documents. Where no current policy covers the question, it says so rather than improvising.
We find out a worker's credential expired when we try to roster them
Credential and training currency is tracked against roster requirements, with expiry flagged in advance. The system warns; the rostering coordinator decides.
We hold sensitive information and cannot say where all of it is
We discover and classify where personal and health information actually sits across your cloud environment, then apply retention, access control and monitoring proportionate to its sensitivity.
We want to use AI but we are not comfortable with the risk
We start with a responsible-AI assessment that maps each candidate use case against privacy obligations, the harm if it is wrong, and whether a human can meaningfully review the output. Several use cases usually fail that test, and we say which.
Evidence from the field
Where this has actually worked, and where it has not.
Published results from other organisations, cited so you can check them. None of these are our clients and none of these numbers are ours.
Where the gains are real
The clearest evidence in any sector, and it is entirely administrative. Ambient documentation gives clinicians time back without touching a clinical decision, which is precisely why it cleared governance and scaled where other health AI did not.
What the evidence does not support
None of this evidence is clinical, and we do not extend it there. Published results also depend on quality assurance running continuously over generated notes, which is an operating cost rather than a one-off implementation.
The Permanente Medical Group and Kaiser Permanente
United States, Northern California
Ambient AI scribes generating draft clinical documentation from the consultation itself, for physician review and editing.
Across 7,260 physicians and roughly 2.5 million patient encounters between October 2023 and December 2024, an estimated 15,791 hours of documentation time saved, equivalent to about 1,794 eight-hour working days.
What it does not prove: Documentation time only. The publication also describes the ongoing quality assurance the organisation had to run alongside it.
Stanford HAI, AI Index
Global
Tracked clinician-reported effects of automatic clinical note generation as adoption scaled through 2025.
Physicians reported up to 83 per cent less time spent writing notes, alongside meaningful reductions in reported burnout.
What it does not prove: Clinician self-report, which tends to overstate against measured time.
Sutter Health
United States, Northern and Central California
Evaluation of an ambient AI documentation platform, using electronic health record audit metrics plus pre and post implementation surveys.
Among 100 clinicians (92 with EHR metrics, 57 completing both surveys), mean time in notes per appointment fell from 6.2 to 5.3 minutes in the 3 months after the April 2024 launch. After hours EHR time did not change significantly, moving from 38.2 to 39.8 minutes. Clinicians reporting they could give patients undivided attention rose from 57.9 per cent to 93.0 per cent.
What it does not prove: Single organisation, purposive sampling of volunteer clinicians, and only 3 months of follow up. The null result on after hours work matters as much as the note time saving: the documentation burden moved, it did not leave the evening.
Mass General Brigham and Emory Healthcare
United States, Massachusetts and Georgia
Survey study of pilot users of ambient documentation technology, measuring self-reported burnout and documentation related wellbeing.
Across more than 1,400 physicians and advanced practice providers (873 at Mass General Brigham, 557 at Emory Healthcare), burnout prevalence fell by 21.2 percentage points at 84 days at Mass General Brigham, and documentation related wellbeing rose by 30.7 percentage points at 60 days at Emory.
What it does not prove: Self-reported survey measures, not measured time or clinical output, and published by the health system about its own study. Response rates were roughly 30 per cent at 42 days and 22 per cent at 84 days at Mass General Brigham and 11 per cent at Emory. The authors state the findings likely represent the experience of more enthusiastic pilot users.
Mass General Brigham (reporting its JAMA Network Open study), 2025
Mass General Brigham
United States, 5 hospitals
Comparison of clinicians using AI scribes against a control group of clinicians who did not, using electronic health record audit log data over more than 2 years.
Across more than 1,800 AI scribe users and 6,770 control clinicians, EHR use fell by 13 minutes a day (a 3 per cent relative decrease) and documentation time by 16 minutes a day (a 10 per cent relative decrease), alongside 0.5 additional patient visits a week and about US$167 a month in additional revenue per clinician. Only 32 per cent of users reached the adoption level (more than half of visits) associated with the larger reductions.
What it does not prove: Observational rather than randomised, and summarised here by the health system that ran it. The authors note the time reductions are too modest to fully account for the improvements in burnout reported in other studies, which is the point worth carrying to a business case.
Gold Coast Hospital and Health Service
Queensland, Australia
A 16 week outpatient trial of an AI-enabled ambient listening scribe across medical officers, registrars and clinical nurses, with blinded scoring of note quality and staff and patient surveys.
Across 100 clinicians and 7,499 consultations between July and December 2024, 58 per cent of AI generated notes were accepted verbatim. Blinded quality scoring put AI notes at 37.06 out of 40 against 34.6 for clinician written notes. 84 per cent of staff reported improved efficiency and 68 per cent of patients said clinicians spent more time in direct conversation. However 47 per cent of staff reported hallucinations, 20 per cent said they occurred frequently, and 16 per cent observed bias in outputs.
What it does not prove: Single site, and the note quality comparison rests on only 18 note pairs. Staff survey response rate was 43 per cent with just 22 patient surveys, the efficiency claims are self-reported rather than timed, and no economic evaluation was done.
Michigan Medicine, external validation of the Epic Sepsis Model
United States, Michigan
Independent external validation of a proprietary sepsis prediction model that had been deployed across hundreds of United States hospitals, run by researchers with no commercial interest in the model.
Across 38,455 hospitalisations among 27,697 patients between 6 December 2018 and 20 October 2019, the model achieved an area under the curve of 0.63 (95 per cent CI 0.62 to 0.64). It failed to identify 1,709 of 2,552 patients with sepsis (67 per cent), while generating alerts on 6,971 of all 38,455 hospitalised patients (18 per cent).
What it does not prove: A predictive model rather than a generative AI system, and a single academic health system. It is included because it is the clearest published example of a widely deployed clinical model performing far below its marketed claims once someone outside the vendor measured it.
Sector figures are useful for deciding what to try first, and almost useless for forecasting your own result. The organisations above had the data, the systems access and the governance in place before they measured anything. The full global picture covers 13 industries, sets out what separated the organisations that got a result from the ones that got nothing, and carries the randomised trials that found no gain at all.
Use cases
What we build in this sector.
Each is tagged with the disciplines involved, because most useful work crosses more than one.
Administrative workflow automation
Automate intake paperwork, service agreement administration, referral triage logistics and internal approvals. Administrative routing only, never clinical prioritisation.
- Automation
- AI
Governed policy and procedure search
Answer procedural questions from the current approved policy set, with the effective date and citation shown, and superseded versions excluded.
- AI
- Security
Workforce and rostering analytics
Credential currency, training compliance, shift coverage and agency reliance modelled against roster requirements.
- Data
Accreditation evidence assembly
Compile policy versions, training records and incident documentation into a reviewable evidence pack for accreditation and funding reporting.
- Automation
- Data
Data discovery and classification
Find where personal and health information sits across the cloud estate, classify it and apply proportionate retention and access controls.
- Security
- Data
Responsible AI assessment
Assess candidate use cases against privacy obligations, potential harm and reviewability, and rule out the ones that should not proceed.
- AI
- Security
Identity and access uplift
Multi-factor authentication, least-privilege access, joiner-mover-leaver automation and access review for a high-turnover workforce.
- Security
- Cloud
Human oversight
Where a person still decides.
Automation proposes. A named person approves. These are the points we design the workflow to stop at.
- A quality or compliance manager approves every accreditation and funding evidence pack before submission. Nothing is submitted automatically.
- The policy owner approves any policy document before it becomes retrievable as current guidance.
- A rostering coordinator approves every shift assignment. Credential and currency checks warn, they never auto-assign.
- A privacy officer approves the scope of any data discovery or classification exercise before it runs, and reviews the findings.
- An executive sponsor approves each AI use case proceeding past assessment, informed by the documented harm and reviewability analysis.
What we do not do
- We do not provide clinical decision support, diagnosis, triage or treatment recommendations, and we do not build systems that produce them.
- We do not build tools that assess, rank or prioritise individuals by clinical need or risk.
- We do not make claims about patient, resident or participant outcomes. Our work is measured on administrative and operational results.
- Where a proposed use case sits close to a clinical decision, we will say so and decline it rather than reframe it as administrative.
- Any system touching health information is designed with the privacy officer involved from the assessment stage, not consulted at the end.
Delivery options
On Azure, on AWS, and afterwards.
The platform is chosen for the workload, not for us. Both paths end in the same place: someone still owns it after go-live.
- 01
Delivered on Microsoft Azure
Suits providers already using Microsoft 365 for policy libraries and staff communication, where governance can be extended from tooling that is already in place.
- Microsoft Purview discovers and classifies personal and health information, and applies retention and sensitivity labelling.
- Azure AI Search indexes the current approved policy set with version filtering that excludes superseded documents.
- Microsoft Foundry hosts the policy assistant with logged interactions and grounding restricted to approved content.
- Microsoft Entra ID and Entra ID Governance deliver multi-factor authentication, least privilege and access reviews for a high-turnover workforce.
- Microsoft Fabric and Power BI provide workforce, credential currency and service delivery reporting.
- Microsoft Defender for Cloud monitors posture across the environment holding sensitive information.
- 02
Delivered on AWS
Suits providers with bespoke client management systems or an existing AWS estate, and those wanting the data platform separate from the productivity suite.
- Amazon Macie discovers and classifies personal and health information across Amazon S3.
- Amazon OpenSearch Service indexes the approved policy set with version-aware filtering.
- Amazon Bedrock runs the policy assistant, with Bedrock Guardrails preventing answers outside the retrieved approved content.
- AWS IAM Identity Center centralises workforce access with least-privilege roles and review cycles.
- AWS Glue and Amazon Redshift build workforce and service delivery reporting, presented through Amazon Quick.
- AWS CloudTrail and AWS Security Hub provide the audit trail and posture monitoring expected of an environment holding sensitive information.
- 03
Operated after handover
Policy sets change, staff turn over and obligations shift. Governance that is not maintained becomes governance on paper only.
- Policy index freshness verified against the document management system so superseded guidance can never be returned as current.
- Access reviews run on the agreed cycle, with high-turnover roles prioritised.
- Data classification re-run periodically as new repositories and systems appear.
- Assistant interactions sampled for accuracy and for any answer straying toward clinical territory, which is treated as a priority defect.
- Security posture findings triaged and remediated with the provider's privacy officer.
- Monthly reporting on administrative time recovered, evidence pack turnaround and outstanding governance actions.
Related services
The capabilities behind this work.
Security and Governance
Close the ways in, know exactly who can do what, and answer an auditor or a client questionnaire from current evidence rather than memory.
Artificial Intelligence
AI that is chosen for a reason, costed before it is built and governed once it is live.
Data and Analytics
Reporting your executives trust, produced once, with a modelled layer underneath it that the next AI or forecasting project can stand on.
Managed Services
Your platform keeps earning its business case after go-live, with cost, security posture, reliability and adoption reviewed on an agreed cycle rather than left to drift.
Free discovery workshop
Start with one healthcare and community challenge.
Bring a process that costs more than it should. We will map the opportunity, the readiness gaps and a recommended next step.