Innovation Analytics

The Innovation Analytics programme will work collaboratively to develop and share innovative projects and best practice in data analytics which could benefit the health and care sector.

Our main areas of interest are as follows: 

  • Synthetic data generation 
  • Natural language processing  
  • Practical applications of advances in machine learning
  • Deploying and productionising machine learning (MLOps)
  • AI safety - explainability, fairness, privacy and robustness in machine learning and AI
  • Other areas, such as Bayesian methods for forecasting

Leading by example, we will code in the open, share our work in the open (via NHSX GitHub), and use open source tools. We are focussed on any innovation work being led by NHSX, but also seek to support teams in the wider NHS attempting to share models and work in interoperable ways.  

There are a number of ways we will support teams:

  • Producing proof of concepts – good concepts are sometimes not taken forward because the project owner hasn’t the time or experience to test out a new approach.
  • Investigating new tools and techniques that can transform the efficiency of existing work - often delivery teams have no spare capacity to investigate such opportunities.
  • Giving technical advice and recommendations, acting as a technical consultancy to support staff leading on projects by offering previous examples and best practice to bolster and build their confidence.
  • Horizon scanning – in a rapidly evolving field it’s hard to know what’s currently available and what’s next. We continually look at emerging software, current best practice and the potential of our underused data.​
  • Offering coding support – the best way to learn coding for data science and analytics (for example Python, R, SQL) is to have practical examples to have a go at and a network of friendly colleagues with whom you can bounce issues off and turn to for advice regarding the best approach.  
  • External collaboration – connecting experts to users through recommendation and advice. 

If you think we can help you, email us

Related Pages