If you have any questions about Real Artificial Intelligence, you can contact the organisers by email on



Real Artificial Intelligence - Speakers

Details of the speakers and their presentations are given below.

Dr Nicola J. Millard (BT)

Botman vs. SuperAgent: how augmented intelligence is changing customer experience.

We are currently witnessing a customer experience revolution. Will Botman or SuperAgent save the day? Dr Nicola Millard shares insights from global research on how consumers are changing their attitudes towards digital service. Are ‘bots going to take over? How do human agents fit in? Are we in a "post PC" era, where the smartphone, voice and screen are our tools for contact? Using a combination of research and practical experience, we will explore the realities of combining human and machine at the customer interface.

Dr Nicola Millard heads up Customer Insight & Futures in BT's Global Services Innovation Team. Despite working for a technology company, Nicola isn't a technologist and combines psychology with futurology to try and anticipate what might be lying around the corner for both customers and organisations (sadly, her crystal ball is broken).

Nicola recently celebrated her 27th year in BT. She has done a number of jobs around the BT business, including user interface design, customer service and business consulting. She was involved with a number of BT "firsts", including the first application of artificial intelligence into BT's call centres and BT's initial experimentation with home working. Nicola got her PhD from Lancaster University in 2005 on motivational technologies in contact centres, published her first book in 2009 and now spends most of her time doing research, writing blogs, articles and white papers.

Nicola regularly pops up on radio and TV around the world, including appearances on 'Woman's Hour', 'Tech Tent', 'The Genius of Invention' and 'Back in Time for the Weekend' for the BBC. She has done 2 TED talks and is also a judge on a number of award panels, including the Institute of Customer Service, the CCMA and the Association of Business Psychology awards. In 2014, she was the recipient of the 'Outstanding Industry Contribution' award from the UK Contact Centre Forum. When she's not doing all that, Nicola travels around the world presenting at conferences and running workshops with an assortment of BT's large multinational corporate clients including banks, government, travel companies and retailers, to name but a few.

Prof. Paulo Lisboa (Liverpool John Moores University)

Good Practice with Machine Intelligence for Real-World Data

Machine intelligence in practice is often synonymous with application of segmentation and predictive models. However, models may lack accepted features of human intelligence such as novelty detection and adaption, for clustering. In classification, they are commonly evaluated on the basis of predictive performance, lacking appropriate consideration of other aspects which can be critical for their use by a subject expert who is not a CI expert. Yet appearances can be deceiving, especially with summary performance measures e.g. AUROC. This is especially the case for non-linear models given their ability to exploit any weaknesses in the data, for instance structural artefacts which can add confounding effects. In addition, many applied CI models work well for well classified cases but cannot explain predictions for borderline cases. In other words, they confirm to expert users what they already know but do not add insights to the data in the difficult cases for which CI is most needed.

There is a drive for the use of CI to complement rather than automate decision making This is fundamental to make CI useful in practice and has been termed Augmented Intelligence, or AI 2.0. The talk will illustrate some of the pitfalls in the design and validation of databased models. It will then describe how rules can be efficiently derived from neural networks so opening the black-box. An alternative and popular way of presenting and using complex models e.g. to clinicians, is the use of nomograms. They will be derived from SVMs so extending this graphical approach to non-linear models. This provides a direct route to interpretation and a way for subject experts to access generic non-linear models constituting good, if not critical and essential practice, for the validation of predictions by machine intelligent models.

Paulo Lisboa is Professor and Head of Department of Applied Mathematics at Liverpool John Moores University, where he is also Head of Engineering and Technology Research Hub and Project Director of LCR Activate, a £5m ERDF funded SME support programme in the Liverpool City Region. He has over 250 refereed publications and his research focus is interpretability of machine learning models for validation by expert users. He has extensive experience in the application of machine learning to clinical decision support, sport analytics and digital marketing.

He chairs the IEEE Computational Intelligence Society Task Force on Medical Data Analysis and co-chairs on Explainable Machine Learning. He is past chair of the H2020 Advisory Group for Health, Demographic Change and Wellbeing, EC and the Healthcare Technologies Professional Network in the Institution of Engineering & Technology, UK. He also has editorial and peer review roles in a number of journals and research funding bodies.

Paulo Lisboa studied mathematical physics at Liverpool University where he took a PhD in theoretical particle physics in 1983. He was appointed chair in Industrial Mathematics at Liverpool John Moores University in 1996 and Head of Graduate School in 2002.

Dr. Federica Sarro (Department of Computer Science, CREST, SSE, University College London)

Living and Working in a Data-driven World: Predictive Analytics for Software Engineering

Software has nowadays pervaded all aspects of our lives. This allows the production and collection of a large amount of information about people’s behaviours and decisions. Predictive analytics exploits such information through intelligent systems able to identify patterns and predict future outcomes and trends. Applied to Software Engineering, it can help us better understand software processes, products and customers in order to maximise product quality, users’ satisfaction, and revenues.

In this talk I will present some of the automated approaches I have devised based on data mining, artificial intelligence, machine and statistical learning, which have proved to be successful to this end. I will also discuss the challenge to deploy predictive models able to turn the large amount of raw data, daily produced by creating and running software, into insights that software engineers can act on, for a wider industrial uptake.

Federica Sarro is a Senior Lecturer (Associate Professor) at University College London (UCL) in the Department of Computer Science. She is also the Director of Appredict Ltd., a software analytics company spin-off of UCL, and member of the CREST research centre.

Her research covers predictive analytics for Software Engineering (SE), Empirical SE and Search-Based SE, with a focus on software effort estimation, software sizing, software testing, and mobile app store analysis. On these topics, she authored more than 60 papers, published in prestigious peer-reviews software engineering conferences and journals. She has also received six international awards, including three best paper awards and the GECCO-HUMIES medal awarded for the human-competitive results achieved by her work on multi-objective effort estimation, and has attained funding for her research in predictive analytics for SE from the Microsoft Azure Research Programme.

She is a very active and highly regarded member of the Software Engineering community: over the last four years she has organised and chaired more than 15 international conferences and workshops and served on more than 50 program committees, and in 2015 she has been elected by the community as a member of the Steering Committee of the International Symposium on Search-Based Software Engineering.

She is also Associate Editor of many prestigious journals such as the Empirical Software Engineering (EMSE) and the IEEE Software (Mobile Applications and Systems area blog), and she has also been Guest Editor for the journals IEEE Transactions on Evolutionary Computation (TEVC) and Elsevier Information and Software Technology (IST).

Dr. Izzy Sargent (Ordnance Survey)

Why OS Needs AI

Ordnance Survey creates and supplies high quality geospatial data for a wide range of customers. These digital representations portray the man-made (e.g. roads, buildings and fields) and natural (e.g. woodland, moorland and boulder fields) characteristics of the landscape for a wide range of applications such as routeing, asset management, planning and geospatial modelling and prediction. Where accuracy, precision and consistency were once primary measures of quality, now increasing emphasis is placed on the capability to generate tailor-made data and solution services to diverse customers. We have set out to develop an approach to data creation that will serve both immediate requirements and pioneer new ways of representing the landscape. In this talk, we will present our work on learning and applying deep landscape representations.

Dr. Isabel Sargent is a Senior Research Scientist at Ordnance Survey, Visiting Researcher at University of Southampton and an industrial advisor for 5 PhD candidates. She specialises in applying machine learning to the interpretation of images of the landscape and particularly excited by discovery from data using unsupervised approaches. Additionally, she is involved in research into the automation of 3D city model capture and quality assessment and has developed operational methods for building height extraction and roof shape classification. Other research interests include the use of crowdsourcing in niche groups and the application of graph analytics in the geospatial domain. She holds a BSc Honours in Environmental Sciences, PhD applying neural networks in oceanography and is a Chartered Scientist.

Ms. Nadia Abouayoub

The Impact of AI in Finance

The purpose of this presentation is to explain how technologies such as AI and RPA (Robotics Process Automation) are currently used in the Financial sector. We will cover their current and futures impacts on the industry. Such impact could be on Business models, business processes and technical architecture. This presentation should be help us to provide a roadmap of where we are and how AI could impact the financial sector.

Nadia Abouayoub is a member of the British Computer Society’s (BCS) Artificial Intelligence Specialist Group Committee (SGAI) and the BCS Women’s Group. She has an MSc. in Computing and DESS in AI and Databases from the University of Dijon, and an MSc. in Formal Methods and the Security of Systems from Royal Holloway, University of London. She has been a member of several committees aiming to empower youth and women, most especially in the finance and new technology sectors. One such initiative saw her organise a VISIO conference with NASA for the World Youth Congress in 2004 to enable participants experience the intricacies of robot navigation and the exploration of Mars. She is the co-organiser of the Machine Intelligence Competition, run by the SGAI and has organised a number of workshops aimed at educating the general public about Artificial Intelligence. She has a personal interest in the United Nations Millennium Development Goals for which she has organised many volunteer workshops. She launched her graduate career with JP Morgan, and has accrued more than 15 years’ experience within the investment banking sector in such fields as Risk Management,Product Control and Technology in Trading Applications. She is currently a Strategist and Improvement Lead.