Meet Our Speakers


Dr Marie Oldfield, CStat, CSci, APAI, SFHEA, FIScT, is the CEO of Oldfield Consultancy and Senior Lecturer at LSE.

Marie is a recognised, published AI and Ethics Expert with a background in Mathematics and Philosophy. She is a trusted advisor to Government, Defence, and the Legal Sector amongst others. Marie also works at the forefront of Ethical AI, driving improvement and development and is the creator of the new AI Professional Accreditation that benchmarks practice in AI and interdisciplinary modelling.

Marie is Founder of the IST Interdisciplinary Artificial Intelligence Group. She was invited to the Executive Board of the Institute of Science & Technology, to be an Expert Fellow for Sprite+ and a member of the College of Peer Reviewers for Rephrain. 

Marie is frequently invited to speak on popular podcasts, panels and at conferences about her experience and research on the development of Ethics in AI and was recently invited to speak at Chatham House on AI and Ethics issues. 

She is passionate about giving back to the global community through extensive pro bono work, with a focus on ethical AI, education, poverty, children and mental health.

 

Talk: Challenges faced by AI Practitioners

What challenges do AI Practitioners face and how can we solve them? Calls for legislation and regulation are accelerating the pace with which we need to get our profession in order. Working with AI and modelling affecting society means you need to not only showcase your skills but prove that you have robust methodology when it comes to your practice, whether you are a philosophy advisor or a computer scientist. 

Marieoldfield

Presenters:

Will Pratt – Will leads the data analytics research and data engineering team at The National Audit Office (NAO). With expertise in large scale data processing and machine learning, Will’s focus is on driving innovation in audit. Before he joined the NAO in 2018, Will worked in the technology sector as a Data Scientist and Software Engineer.

Hisham Taj – Hisham leads the horizon scanning approach from within the modelling network at NAO. Having worked as a machine learning researcher prior to joining the NAO, Hisham has expertise in software deployment and deep learning.

 

Talk: How much AI is in Government – and how is it audited?

The session will open with an introduction to the different types of AI in government, as taken from our research in the VFM paper “Use of artificial intelligence in government.” This session will then primarily focus on audit approaches for machine learning models, drawing on research undertaken with other public audit bodies internationally. We hope to stimulate and facilitate discussion about technical methods in algorithmic evaluation, as well as less technical concepts associated with evaluating whether machine learning methods/tools have been appropriately deployed into public service delivery.

Will Pratt Hisham Taj

Dr Richard Wood

Richard is Head of Modelling and Analytics at NHS Bristol, North Somerset and South Gloucestershire Integrated Care Board. He is also Senior Visiting Research Fellow, University of Bath School of Management.

Talk Title: From data to insight to action: How the NHS is (slowly but surely) catching up

Over the last two decades, there has been a vast improvement in data collected by the NHS. Where information was once kept on paper, it is now stored electronically. Where data was once at aggregate level, it is now of the finer granularity required for richer analysis and linkage at person-level.

While work continues apace – expanding data coverage, frequency and quality – the question must be asked: are we really making the most of this resource, and so justifying its ever-increasing volume? This was a challenge at least partially advanced by Martin Bardsley in his 2017 Health Foundation report, titled “Do we have more data than insight?”. But what’s the point in having insight if we can’t turn it into action? In this talk, I will review a number of analytical projects at my own healthcare system in around Bristol, covering how – through novel application of various quantitative techniques, from random forests to traditional queueing models – we have derived a breadth of insight from a multitude of data sources.

Concluding, I will reflect on the extent to which these efforts have led to demonstrable positive action, and attempt to outline some critical success factors for implementing analytics in practice.

Richard Wood

Anne Liret

Title of the workshop: The difficult task of building fit-for-user explanations for AI models in business

Anne Liret is currently research manager in Operational Transformation and Artificial Intelligence in BT’s Applied Research. Her research for 27 years, includes problem modelling and solving using AI and Optimisation techniques, with a focus on sustainable resource scheduling problems in dynamic context, interactive explainable AI-based decision-making, context-aware self-learning models for real-time diagnosis. Her models have helped transforming operations in field engineering services, asset maintenance and service assurance. She has PhD in AI and formal computation, and a strong experience in applying AI, to transform telecommunications services operations.

Since 2020, she has been studying approaches for reusing explanation experiences, recommending explanation strategies that suit the needs of users, and improving understandings about how explanation is perceived by human-being. She is a member of the European Chistera project called iSee addressing this topic and building an ethical responsibility framework for systems and human-being in “XAI” experience.

Anneliret BT[87]