The evidence behind CompliMind

Measured improvements in time, accuracy, and compliance assurance - validated in NHS settings.

Every claim on this page is traceable to its source. Where evidence comes from controlled experiments, we report sample sizes, effect sizes, and statistical significance. Where data is self-reported, we say so. We do not present pilot outcomes as production-scale results.

What we have measured in practice

What we've measured in practice

150%
Increase in error detection in policy review
Randomised controlled trial, n=33
96%
Accuracy on NHS compliance questions
Independently scored, n=250
400+
Hours/month redirected to forward planning
Somerset NHS Foundation Trust
£11k
Annual savings per senior staff member
CIBSE Paper, Band 8+

What changes in practice

Task
Before
With CompliMind
Policy review
4-7 hrs manual cross-checking
30-60 mins structured review
Finding HTM requirement
10-20 mins searching PDFs
Instant answer with citation
Meeting minutes
Manual drafting
Auto-generated with actions

The compliance challenge facing NHS estates

NHS estates and facilities teams manage some of the most complex regulatory environments in the country. The evidence shows a widening gap between what is demanded and what teams have the capacity to deliver.

Compliance time

The compliance burden

24.5 hrs/week

spent on regulatory searches, policy writing, and compliance reviews - 65% of contracted time.

24.5 hrs/wk
10.9 hrs searching
8.1 hrs writing
5.5 hrs reviewing
leaves only
13 hrs
per week for on-site
maintenance, inspections
& team management
35% of contracted time

CompliMind CRM survey, n=482 | 37.5 contracted hrs − 24.5 compliance hrs

Legal & patient impact

HSE enforcement action

£11.4M

in HSE fines to NHS Trusts (2015–2024). Average fine: £285k per case, with legal costs typically doubling that.

HSE prosecution database

Patient safety impact

5,348

clinical incidents linked to estates or facilities failure in a single year (2021–22). 43% of doctors say premises harm patient care.

NHS NRLS / BMA survey

What NHS teams are achieving with CompliMind

CompliMind is deployed in NHS Trusts across England and Wales. Here is what practitioners and their organisations report from live, sustained deployments.

Somerset NHSFT Case Study Video

Practitioner Testimonial

2–3 minute video from a named NHS practitioner.
Recommended: Paul Luxton (Somerset) or Craig West (Cambridge)

UHSUniversity Hospital Southampton NHS FT

How CompliMind uncovered a long-standing contract inconsistency

£48k
Annual contract savings confirmed
8 years
Time the inconsistency had gone undetected
1 review
Required to identify cost recovery

Southampton NHS FT manages a complex multi-service contract with a large, global facilities management provider. CompliMind set up role-specific Specialists to review the contract, identifying two figures that should both be adjusted for inflation but were being treated differently. The Trust identified £48k in annual savings, with potential for significant retrospective recovery.

"CompliMind found an inconsistency that had been missed for years and built a case for correcting it. Those cost savings can now go to where they're needed most."
Bertie Young
Associate Director of Facilities
📄 Download case study (PDF)
Swansea BaySwansea Bay University Health Board

How CompliMind empowers PFI Handback Teams

6,000+
Asset survey images reviewed
35%
Time savings for the PFI team
£100k+
Potential cost recovery

Swansea Bay is preparing for a PFI handback due in late 2029. CompliMind was introduced to centralise contract access and streamline document analysis for the handback process. Review cycles became faster, communication across estates, facilities, capital projects and contractors became more aligned, reducing risks and delays.

"CompliMind cuts hundreds of hours of manual work. Its speed and consistent analysis let us focus on important decision-making rather than reviewing contract documentation."
Rob Neal
PFI Handback Senior Project Manager
📄 Download case study (PDF)
UHSUniversity Hospital Southampton NHS FT

How CompliMind transformed risk assessment production

280
RAMS produced in under one month
Faster than slow, manual processes
100%
Of required RAMS now up to date

Risk assessments are required before any maintenance task can begin. CompliMind Chat was used to generate structured RAMS at scale, drawing in relevant NHS-specific regulatory references automatically. The team produced approximately 280 RAMS around five times faster than manual processes, with time saved reinvested into day-to-day operations.

"Writing RAMS used to take up a huge part of my week. CompliMind pulls in the regulations automatically and gives me a solid structure to work from. The time it saves is significant."
Rui Rodrigues
Mechanical Team Leader
📄 Download case study (PDF)
IHEEMIHEEM

How CompliMind streamlines meeting documentation across technical platforms

40+
Technical platform meetings per year
2 weeks
Time savings directed to other projects
75
Technical platform members supported

IHEEM, the UK's professional body for healthcare engineering and estate management, relies on a network of technical platforms covering fire safety, ventilation, medical devices, and decontamination. CompliMind's AI Writing tool produces structured minutes from transcript alone, replacing a manual bottleneck. Minutes are now produced, approved, and distributed the same day.

"I run every technical platform meeting through CompliMind now. It identifies the key discussion points, cuts through the waffle, and gives me something I'm confident sending straight to the Chair for approval."
Becky Carter
Committee Business Manager, IHEEM
📄 Download case study (PDF)
Somerset NHS FTSomerset NHS Foundation Trust

400+ hours/month redirected to forward planning

60
Registered users
1,500
Regulatory queries per month
400+
Hours/month redirected

60 users across Somerset FT handling ~1,500 regulatory queries and ~70 sets of meeting minutes per month. Previously reliant on disconnected folders, spreadsheets, and emails. CompliMind centralised compliance knowledge and freed capacity for forward planning.

"CompliMind is like having a compliance colleague with encyclopaedic memory, drawing together regulations that normally sit in isolation."
Paul Luxton
Head of Acute Estates & Infrastructure, Somerset NHS FT
📄 Download case study (PDF)

What is compliance paperwork costing your team?

Based on NHS research across estates and facilities teams at Band 5–9. Input your team size to see evidence-backed savings scenarios.

Staff across Bands 5–9 involved in compliance, assurance, or regulatory work
40
people
5255075100125150
-
spent on compliance paperwork every year
-
Total hours per year on compliance admin
-
Hours per person per week
-
Of working time spent on paperwork
-
Left for on-site work, inspections & management
8.1 hrs
5.5 hrs
Searching for information Writing compliance docs Reviewing & checking

If you reduced that burden

Evidence-backed scenarios
Scenario
Hours redirected
Capacity released
Per person / week

How we handle evidence

All statistics are traceable to source documents. We report sample sizes, effect sizes, and statistical significance. Self-reported data is labelled as such. Pilot outcomes are not presented as production-scale results. Questions about any claim? evidence@complimind.co.uk

Public AI tools are already creating legal and governance exposure

NHS teams are turning to consumer AI tools to cope with compliance workloads. The legal system is catching up faster than most organisations expect - and the consequences are real.

>50%
of employees admit to using AI tools without telling their managers.
Microsoft & LinkedIn, Work Trend Index 2024 →
77%
of employees have pasted sensitive company data directly into generative AI prompts.
LayerX, Enterprise AI Security Report 2025 →
70%
of employees have never received formal training on how to use AI safely or ethically.
Asana/KPMG, State of AI at Work 2024–2025 →

Uploading sensitive data into public AI is now a legal liability

In Munir v Secretary of State [2026] UKUT 81 (IAC), the Upper Tribunal ruled that uploading confidential documents into an open-source AI tool such as ChatGPT places that information into the public domain - constituting a breach of client confidentiality and a waiver of legal professional privilege. Courts now distinguish between open-source and closed-source AI, and will hold organisations accountable for how staff use AI with sensitive data. NHS Trusts using uncontrolled public AI tools for compliance work face exposure to confidentiality breaches, privilege waiver, and regulatory sanction. Adopting a purpose-built, closed-source platform is a governance decision, not just a productivity one.

Government evaluation

M365 Copilot Evaluation - Dept for Business and Trade (Oct 2024 – Mar 2025)

A 1,000-licence pilot found no evidence that time savings led to improved productivity. Control group colleagues had not observed productivity improvements. The evaluation highlighted inconsistencies in quality-assuring AI outputs, hallucinations throughout the pilot, and a lack of knowledge regarding capabilities, limitations, and security concerns among non-pilot staff.

100% 75% 50% 25% 0%
36%
Copilot
74%
ChatGPT
96%
CompliMind

Based on independent scoring against verified NHS Estates & Facilities Management standards (n=250).

CompliMind is built for infrastructure compliance.

  • Sector-specific knowledge base: Always up-to-date with national guidelines and seamlessly integrated with your local document database.
  • Built-in domain expertise: Built on a deep understanding of Capital, Estates and Facilities, with best-practice templates that align guidance with compliance and assurance workflows.
  • Built for auditability: Every answer is linked to its regulatory source, providing transparent, version-controlled, and fully auditable information.
  • Privacy and security: Designed to protect Trust data through enterprise-grade compliance.

The research behind the platform

Controlled NHS studies and independent benchmarking.

CompliMind research presented to NHS estates and facilities professionals
Randomised Controlled Trial

AI-Assisted Policy Review

33 NHS EFM professionals. Manual review vs. CompliMind-assisted review of a policy derived from HTM 07-01. Statistically significant gains in error detection - no increase in false positives.

  • 150% increase in error detection recall (p=0.003)
  • 110% improvement in detection efficiency
  • overall detection quality (F1 score)
  • r > 0.60 very large effect sizes across all metrics
📄 Download summary (PDF)
Controlled Experiment

Information Retrieval (n=64)

64 NHS EFM staff tested on compliance tasks. Quality scored against verified standards. Triangulated by three independent researchers. Experienced staff benefit most.

  • 35% faster task completion
  • 53% improvement in answer quality
  • 117% quality gain for 30+ years' experience
  • £3.3–11k annual saving per staff member
📄 Download CIBSE paper
Independent Benchmark

Platform Accuracy (n=250)

250 verified NHS EFM standards (HTMs, HBNs, HSE). CompliMind vs. generic AI. Closed-source architecture avoids legal privilege risks confirmed by the Upper Tribunal (Munir, 2026).

  • 96% accuracy on compliance questions
  • accuracy vs manual reviews
  • Closed source - no public domain data risk
📄 View methodology
Key Insight

This is not about saving time. It's about reducing risk.

Experienced professionals benefit the most. Error detection doubles. Every answer is traceable to its regulatory source. When procurement teams ask "does this save time?", the better question is:

"What is the cost of the errors it catches?"

  • £11.4M HSE fines to NHS Trusts (2015–2024)
  • 5,348 clinical incidents from estates failure
  • £20M maximum exposure (Corporate Manslaughter Act)
CompliMind team at IHEEM Healthcare Estates Conference

Publications & Professional Endorsement

📰

IHEEM Health Estate Journal

October 2025. Peer-reviewed. Compliance burden and AI-assisted tooling in NHS estates.

Read article →
🎓

CIBSE Technical Symposium

Controlled experiment (n=64). Productivity by salary band.

Read abstract →
📋

CIBSE Journal

Cambridge research programme. Hallucination prevention approach.

Read article →
🔬

IFHE / DRLC Article

External AE endorsement. Dr Louise Webb, DRLC Ltd.

Read article →
ISO
27001
Information Security
ISO
42001
AI Management
ISO
9001
Quality
ISO
14001
Environmental
DTAC
NHS Digital Tech
GDPR
Data Protection
CE+
Conformity
University of Cambridge
NHS England
IHEEM
CIBSE
NPAG

See the evidence in your context

Every Trust is different. We can show you how the evidence applies to your specific compliance burden - and demonstrate the platform on your own documents.