Use Case Intelligence RSS Search use cases Search
KPMG Healthcare Denmark

Radiometer

Enterprise Knowledge Search

Radiometer wanted a responsible way to design, build, and deploy an internal AI system that could generate exact insights, support scientific advancement, and help employees across knowledge, sales, service, and recruiting work without compromising ethics, governance, or security.

Read the original story Explore all use cases
around 4,800 employees worldwide
over 130 countries/jurisdictions

Reported impact

Business outcome

Radiometer deployed RADChat across multiple internal functions, giving employees a governed generative AI assistant that automates language-based work with greater speed and efficiency while helping the company build trust in responsible AI and refine additional use cases.

  • Productivity Improvement

Solution

Radiometer worked with KPMG in Denmark to design and deploy RADChat, a proprietary internal generative AI system hosted on Microsoft Azure. The implementation started with an AI roadmap, safe-usage guidelines, and a requirements backlog for a data-integrated LLM environment so the team could govern how enterprise knowledge and business content would be used. Microsoft Azure provided the cloud platform for the business-critical application, while RADChat exposed the model to employees through use cases spanning knowledge management and retrieval, customer sales and service operations, and recruitment and onboarding. KPMG's governance frameworks, controls, processes, and tools shaped the ethics, security, and operating model around the rollout so employees could experiment with the assistant, help evolve it, and use it as a trusted productivity tool rather than a disconnected proof of concept.

Data flow

Internal business and knowledge content is integrated into the Azure-hosted LLM environment, RADChat generates answers and language-task outputs from that governed context, and employees consume the responses in knowledge retrieval, customer sales and service, and recruitment and onboarding workflows.

Solution architecture

4 components · 3 layers
  1. Compute
    • Microsoft Azure Hosts the internal AI system on the cloud platform Radiometer chose for business-critical applications.
    • data-integrated LLM solution Connects enterprise knowledge and business content into the large-language-model environment used by employees.
  2. Serving
    • RADChat Serves as the employee-facing assistant for knowledge retrieval, customer operations, and recruitment workflows.
  3. Governance
    • governance frameworks, controls, processes and tools Provides the ethics, governance, security, and safe-usage guardrails around the internal AI deployment.

Architecture clues

  • Deployment planning included an AI roadmap, safe-usage guidelines, and a requirements backlog for a data-integrated LLM solution.
  • Microsoft Azure was selected because Radiometer already relied on Microsoft for business-critical applications.
  • Radiometer and KPMG co-developed RADChat as a proprietary internal LLM assistant.

Evidence from the source

Creating a backlog of the requirements needed to deploy an internal, data-integrated LLM solution was a key part of Radiometer's collaboration with KPMG in Denmark, along with defining the organization's AI roadmap, and establishing safe usage guidelines.
RADChat, the proprietary LLM AI system co-developed by KPMG professionals and Radiometer, and powered by Microsoft Azure, has been deployed for use across a range of applications within the organization, including knowledge management and retrieval; customer sales, services and operations; and recruitment and onboarding.
Radiometer wanted to explore the potential of artificial intelligence (AI) as a means of generating exact insights and fueling scientific advancements.
Radiometer was looking to find a responsible approach to designing, building and deploying an internal AI system.
The tool is already transforming the business by automating and executing language-based tasks with unprecedented speed and efficiency.
Transparency Evidence & methodology 71/100 reference value · 83% confidence · 1 source

Review how this reference was scored and trace each normalized claim back to its documented source.

01

Scoring methodology

How scores are calculated

Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.

Customer reference value71/100
  • Problem clarity20/20
    Measures whether the customer's operational problem is specific enough to compare.
  • Implementation depth25/25
    Rewards named components, data flow, integrations, and architecture detail.
  • Outcome specificity16/20
    Measures whether the reported benefit is concrete rather than a generic claim.
  • Quantified impact0/25
    0 metrics describe measurable impact.
  • Deployment scale10/10
    Rewards evidence of a live or scaled deployment and meaningful integration scope.
Evidence confidence83%
  • Source authority20/25
    The main evidence is an official vendor-published customer story.
  • Claim traceability25/25
    12 of 12 key fields link to a named source.
  • Direct evidence20/20
    5 direct source quotes retained.
  • Date provenance0/10
    The publication date has no documented provenance.
  • Source transparency10/10
    1 source reference document publisher, URL, authority, retrieval, and supported fields.
  • Implementation specificity8/10
    Rewards source-backed components, data flow, quantified impact, and a specific outcome.
02

Traceability

Sources & provenance

Field coverage shows which facts each source supports. Dates are shown only when documented by the source.

  1. KPMG

    Unlocking the power of generative AI at Radiometer

    First-party
    https://kpmg.com/xx/en/client-stories/unlocking-the-power-of-generative-ai.html
    Source type
    Vendor-published customer story
    Supports
    • Customer
    • Industry
    • Region
    • Company size
    • Business problem
    • Solution
    • Products
    • Solution components
    • Data flow
    • Integrations
    • Business outcome
    • Metrics
    • Architecture
    • evidence quotes

Date evidence

  • First archived Mar 7, 2026 · Internet Archive · Independent First Wayback Machine capture; confirms the page existed by this date, not when it was published.
    Open date source
  • Sitemap last modified Mar 31, 2026 · KPMG · Publisher metadata Publisher-reported modification timestamp; not a publication date.
    Open date source

Related

Similar use cases

See all like this
Databricks Oct 8, 2026 ModMed Healthcare United States ModMed needed to protect sensitive patient data and maintain compliance while reducing administrative overhead, improving access to operational data, and creating AI-enabled products and agentic workflows for ambulatory… Enterprise Knowledge SearchReal-time Analytics Databricks Data + AI Platform · Genie Agents · Genie Code Databricks Sep 11, 2026 Helix Healthcare United States Helix needed to replace a fragmented research environment and niche tooling that made it difficult to search petabytes of genomic and patient data efficiently and securely across teams. petabytes of clinical genomic data Enterprise Knowledge SearchSecurity and Compliance Databricks Data + AI Platform · Genie · Unity Catalog ServiceNow Modified · Jul 31, 2026 Wellstar Health System Healthcare Georgia, United States Wellstar clinicians needed help inside the tools they already used, but disconnected service systems forced staff to hunt for answers, slowing care and making growth by merger hard to absorb without new support hires. 11 FTE capacity created through automation Enterprise Knowledge SearchIT Operations Automation IT Service Management · Moveworks · Now Assist for ITSM