Working with the most ambitious projects...
..... and the worlds largest employer

We have worked with a number of organisations and clients, all with their unique requirements. Please read our case studies below to see how we helped them overcome challenges and improve their data use.

Case Studies

NHS Digital

Challenges they were facing:

Hundreds of separate data collections (with over 2000 permutation based on local data flows)
Datasets in a variety of formats: xslx, csv and xml
Data quality problems
Maintenance of current data pipeline expensive

How we helped:

Complete Data Dictionary Import including business definitions, data types, attributes and machine processible business rules
Training for key team members
Cloud-hosting, installation, and support for the Metadata Exchange and Metadata Monitor
Reusable data pipeline to validate, store and report on data using Regex, Drools (DRL) and DMN
Test harness to automate data quality process based on reusable rules and constraints

North Thames GMC


Complex data, multiple events, data drawn from 100’s of different systems and
Collected data items meaningless as not translated into a one common
terminology and sparse metadata.
Data is coded within different hospitals in many formats.

How we helped

Worked with local teams from seven Trusts to document local pathology formats.
We ran Elastic search matching over their datasets to “suggest terms” and gave these to clinical teams to validate
Created Mappings for locally coded pathology results to standard so it could be queried consistently going forward
Shared information within the Metadata Exchange platform that sits in their Open EHR architecture

100K Genomes Project


Information initially collected via spreadsheets.
Data difficult to find.
Unstandardised data definitions, limited documentation and silos.

How we helped:

Deployed the Metadata Exchange, which acts as a knowledgebase for standards and specifications that describe the data collected and stored by the organisation.

Generated Digital Forms which only have options for data entry as the model catalogue specifies so that no deviation of data format can cause irregularities.

Significantly reduced in time to specify datasets and data officers can now spend their time on more important activities such as defining correct formats for data entry by the clinical researchers.