Skip to content

Latest commit

 

History

History
75 lines (61 loc) · 5.39 KB

assembly.md

File metadata and controls

75 lines (61 loc) · 5.39 KB
Raw
layout title
default
Data Science Assembly

Data Science Assembly

What is it?

A group that represents those working in novel data science (both R&D and operational) in NHSE/I* (inc NHSTD* and NHSD*). For example, those that are involved in blue-sky research and trialling novel approaches to data science and analytics, rather than performing operational analytics with an immediate business need. We are not proposing to alter existing day-to-day working processes, rather be a group that can represent the requirements and interests of its data scientists. Our role would include (but not be limited to) things like:

  • A forum to collect requirements and views
  • Educate others, advocate for data science

Why do we need it?

  1. Currently R&D analytics and data science occurs in pockets throughout NHSE/I (for example, the NHSTD AI skunkworks, the innovative analytics team with the NHSTD Analytics Unit and the data science team in NHSE/I). This is in contrast to the more centralised processes for operational analytics that occur within, for example, the CDAO directorate in NHSE/I.
  2. Currently, the platforms are not set-up to allow easy NHSE/I-NHSTD collaboration (primarily flow of data and codes but also tooling, etc.). We need a better understanding of if/how folks currently collaborate and a roadmap to enable easy (but secure) collaboration.

However, representation from a strong and recognised Data Science function is important, to ensure novel approaches and techniques can feed into operational analytics, AI and beyond. In addition to experimenting/filtering R&D projects and recommending successful ones to operational analytics, we can provide important guidance/reference to the operational team in terms of the latest tech (e.g. algorithms/approaches, tools, tech stack, etc.)

Pharmaceutical Industry - an example, for illustrative purposes:

The pharmaceutical industry, like many industries, relies on its R&D laboratories to innovate, experiment, try new things and make new discoveries. Many (most!) of these experiments do not work. However, hits can then be passed into production, using industrialised processes in factories and large manufacturing facilities. Here, in these factories, consistency, control and repeatable processes are key. However, the needs of an R&D lab are different; here flexibility and the ability to try new things (and quickly move on to the next thing) are key. The R&D lab needs different equipment that you wouldn't find in the factory. So, if you only had the factory, or you tried to do all your R&D in the factory, you are unlikely to get very far. The same is true with R&D analytics and data science.

Currently, we may get asked individually, or not at all, to contribute to the development of: Data Science projects

  • Analytics tools, software and platforms
  • Workforce, for example:
    • Training and development for existing colleagues
    • Workforce planning, e.g. attraction and retention

A data science assembly would allow us to collaboratively contribute to these developments with a stronger and more consistent voice, to ensure the requirements of data science and R&D are an appropriate part of future plans. We can also share the load between us.

Who will be in it?

Members will be from teams contracted to work for NHSE/I, with enthusiasm and experience in Data Science and innovative analytics. This includes:

  • Giuseppe Sollazzo - AI Skunkworks, NHSTD, NHSE/I
  • Jennifer Hall - AI Imaging, NHSTD, NHSE/I
  • Matthew Cooper - AI Skunkworks, NHSTD, NHSE/I
  • Amadeus Stevenson - AI Skunkworks, NHSTD, NHSE/I
  • Ed Kendall - Data Science Hub, CDAO, NHSE/I
  • Achut Manandhar - Data Science Hub, CDAO, NHSE/I
  • Svetlana Batrakova - Data Science Hub,CDAO, NHSE/I
  • Chris Mainey - Patient Safety, NHSE/I
  • Sarah Culkin - Analytics Unit, NHSTD, NHSE/I
  • Jonny Pearson - Analytics Unit, NHSTD, NHSE/I
  • Dan Schofield - Analytics Unit, NHSTD, NHSE/I
  • Paul Carroll - Analytics Unit, NHSTD, NHSE/I
  • Rupert Chaplain - Data Science Team, NHSD
  • Jonathan Hope - Data Science Team, NHSD
  • Mohammed Absar - Direct Commissioning & Medical Directorate - London Region, NHSE/I

What will it do?

Collectively represent the R&D analytics and data science community in NHSE/I (including NHSTD & NHSD), including:

  • Develop tech stack needs; open source, security, flexibility for R&R, separate to, but interoperable with operational platform (e.g. Foundry)
  • IG challenges going from R&D to device
  • Role in education others such as IG on what AI is going
  • Be a centre of excellence
  • Working together with the Data Science Hub to develop capability across NHSE/I
  • Share our work being undertaken in the R&D space

How will it do this?

  • Communication of our existence and role to key stakeholders
  • Build collective links to Data Engineering/ Data Services
  • Our representatives can share the attendance of key stakeholder meetings
  • Group meets regularly (quarterly?) to feedback and decide on our collective actions/positions
  • Dissemination of our work e.g. through publication, store of case studies, annual report

What is it not?

  • A formal organisation with secretariat and admin function
  • A team from which projects and data science outputs are commissioned (this would still happen in separate teams)
*Acronym Expansion
NHSE/I NHS England & Improvement
NHSTD NHS Transformation Directorate, formerly NHSX
NHSD NHS Digital