DATA SCIENCE, AI, ML, DEEP LEARNING, BEAUTIFUL ENVIRONMENTS, URBAN DESIGN, HUMAN WELLBEING
Chanuki Illushka Seresinhe
Passionate about pushing the boundaries of technology to create a positive global impact.
Biography Invited Talks & Presentations Press & Videos Publications & Policy Briefs Professional Experience Contact me
BIOGRAPHY
Dr. Chanuki Illushka Seresinhe is a leading figure in the data science community, known for her impactful contributions across both commercial sectors and non-profit initiatives.
In her most recent role as Head of Data Science at Zoopla and Hometrack, Dr. Chanuki Illushka Seresinhe spearheaded the integration of machine learning to refine the user experience for those looking to buy, sell, or rent properties. Her forward-thinking initiatives, such as the application of generative AI for area guides and the introduction of NLP and Computer Vision driven smart property tags, have set new benchmarks in property search in the UK. Concurrently, at Hometrack, she led the modernisation of the Automated Valuation Model (AVM), including the incorporation of machine learning to enhance property valuations, marking a significant shift in how property values are assessed in the UK.
Chanuki's commercial experience was first established during her tenure as a commercial data scientist at Channel 4, which laid the groundwork for her subsequent role as the Director of Data Science at Culture Trip. In these capacities, she demonstrated an exceptional ability to translate complex datasets into actionable insights, significantly improving user engagement and business performance through innovative data-driven strategies.
Alongside her achievements in the commercial sector, Chanuki co-founded Beautifulplaces.ai, a non-profit initiative born from her deep interest in the nexus between environmental aesthetics and human well-being. This project extends her doctoral research conducted at the University of Warwick and in partnership with the Alan Turing Institute. There, she delved into the relationship between the visual appeal of our environments and well-being, employing large online datasets and advanced deep learning techniques. Beautifulplaces.ai embodies Chanuki's dedication to societal benefit, demonstrating her ability to leverage data science for a broad spectrum of applications, ranging from solving business problems to improving public spaces.
Her research has received international recognition, with findings highlighted in top publications including the Economist, Wired, The Times, Newsweek, and the BBC, cementing her status as a thought-leader in her field. Chanuki is an esteemed speaker, invited to discuss her work at conferences around the world. This includes prestigious events such as the House of Beautiful Business, the Generative AI Summit in London, YOW! in Melbourne, Brisbane, Sydney, the World AI Summit in Amsterdam, The Turing Presents: AI UK, and others, further contributing to the discourse on the vital role of data science in improving human experience.
Chanuki also belongs to Azeem Azhar’s Exponential Do, the invite-only community of Exponential View that brings together professionals to think, build and invest in the Exponential Age. Here she co-hosts the human embodiment channel along with Robbie Stamp and Luna Lacey.
Before embarking on her data science career, Chanuki had a successful career in the digital domain, running her own digital design consultancy in London for over eight years. Her consultancy provided strategic advice to clients such as the Tate and the Design Museum on maximizing their digital presence. With professional experience spanning across California, New York, and Paris, she has cultivated a rich understanding of digital design's impact on user engagement and business strategy.
Chanuki’s career is a testament to her multifaceted expertise, from leading commercial data science initiatives at Zoopla to pioneering research through Beautifulplaces.ai Her endeavours exemplify how data science can be leveraged not just for business innovation but also as a powerful tool for societal benefit.
Invited Talks & Presentations
House of Beautiful Business - The Fest, Lisbon, September, 2024
House of Beautiful Business - The Gathering, Tangier Morocco, MaY 2024
Generative AI Summit, London, May 2024
Generative AI Marketing Summit, London, February 2024
YOW! 2023 Melbourne, Brisbane, Sydney, November 2023
World AI Summit in Amsterdam, October 2023
Festival Of place, July 2023
Generative AI Summit, May 2023
J on the Beach, Malaga, May 2023
The Rise of Generative AI, Reuters Events, March 2023
Every Woman in Tech Forum, March 2023
Beauty: Who Cares, Wins! Joint conference with RIBA and THE Office for Place, February 2023
Women of Silicon Roundabout Conference, November 2022
Bright Ideas Gathering, OCTOBER 2022
Re-Work Deep Learning Summit, HARNESSING THE POWER OF DEEP LEARNING, SEPTEMBER 2022
AI and YOU Podcast, AUGUST 2022
Women of Silicon Roundabout Conference, November 2021
The AI Summit, September 2021
The Turing Presents: AI UK, March 2021
Social science FOO Camp, Facebook HQ, February 2020
Beyond Conference, R&D conference for the creative industries, Edinburgh, 2019
Data for good, Behavioural Exchange, SEPTEMBER 2019
The Festival of Place, LONDON, JULY 2019
RE-WORK Women in AI, LONDON, JUNE 2019
What can machine learning reveal about beautiful places?, LONDON, MAY 2019
Creating Communities Create Streets Conference The importance of Good design, LONDON, March 2019
Data Matters at Skillsmatter LONDON, March 2019
Ada Lovelace Day Live! (international celebration of the achievements of women in STEM) LONDON, OCTOBER 2018
LSE Centre for Economic Performance Wellbeing Seminar Series, LONDON, OCTOBER 2018
Keynote Speaker, Infinit Conf 2018 LONDON, JULY 2018
Behavioural Science Summit 2018, University of Warwick, June 2018
COGX18, the Alan Turing Institute research stage, London, JuNE 2018
SoapBox Science (public outreach platform for promoting women scientists) LONDON, MAY 2018
YOW! 2017 Melbourne, Brisbane, Sydney, November 2017
Open Data Science Conference London, October 2017
Keynote Speaker, InfiniteConf 2017 London, July 2017
COGX17 - AI and Mental Health Panel, London, July 2017
OpenTech 2017 London, June 2017
Data Beers London, June 2017
The Alan Turing Institute Short talk, April 2017
Warwick Medical School University of Warwick, January 2017
Data Beers Warwick University of Warwick, December 2016
Science & Technology Select Committee visit, Alan Turing Institute London, November 2016
Data Science and Government Conference Blavatnik School of Government, University of Oxford, June 2016
Future Cities Forum, RIBA London, June 2016
PRESS & VIDEOS
Google Public Policy Working together to build AI responsibly
BBC Click Algorithm learns to understand natural beauty
The Economist Computer analysis of what is scenic may help town planners
Wired An AI has been trained to understand beauty
The Times Beautiful cities hold the key to happiness
Wired Why Google Maps and Citymapper are terrible for walking directions
Quartz What two centuries of book publishing tell us about the “good old days”
Newsweek: This was the lowest point in U.S. history in terms of the nation's happiness.
Telegraph Britain 'became happier' after Margaret Thatcher became Prime Minister
MIT Technology Review Were people really happier in the past? Millions of pieces of text suggest not
The Times Victorian times were happiest, study of the national mood finds
Scientific American Scenic City Sights Linked to Higher Happiness
Nesta Reoport Street style: machine learning takes to the streets
The Times City vistas prove easier on eye than green belt
Scientific American 60-Second Science Teaching Computers to Enjoy the View
Microsoft Accelerating data science at the Alan Turing Institute with Microsoft Azure
Guardian Beauty spot or landscape blot? Computer trained to judge scenery
Scientific American 60-Second Science: Photos Tagged as Art Linked to Rising Property Prices
Guardian Does a pretty office make a productive workforce?
Interviewed on Channel 4 Phil Spencer's Perfect Home: "Home Office" episode
PUBLICATIONS & POLICY BRIEFS
Using Deep Learning to Quantify the Beauty of Outdoor Places
Seresinhe, C. I., Preis, T., & Moat, H. S. (2017) Royal Society Open Science, 4(7), 170170.
Is Happiness Greater in More Scenic Locations? Large Scale Evidence From Mobile Phone And Online Data.
Seresinhe, C. I., Preis, T., Mackerron, G. & Moat, H. S. (2019) Scientific Reports, 9, 4498
Street-Frontage-Net: urban image classification using deep convolutional neural networks
Law, S., Seresinhe, C. I., Shen, Y. & Gutierrez-Roig, M. (2018). Street-Frontage-Net: urban image classification using deep convolutional neural networks. International Journal of Geographical Information Science, 1-27.
Historical Analysis of National Subjective Wellbeing using Millions of Digitized Books
Hills, T., Seresinhe, C. I., Proto, E., Sgroi, D. (2019). Nature Human Behaviour, Volume 3, 1271–1275(2019)
An application of convolutional neural network in street image classification: the case study of london.
Law, S., Shen, Y. and Seresinhe, C., (2017) In Proc. of the 1st Workshop on Artificial Intelligence and Deep Learning for Geographic Knowledge Discovery, Redondo Beach, CA, 07 – 10 October 2017, 5-9.
Quantifying the Impact of Scenic Environments on Health
Seresinhe, C. I., Preis, T., & Moat, H. S. (2015) Scientific Reports, 5, 16899.
Quantifying the Link Between Art and Property Prices in Urban Neighbourhoods
Seresinhe, C. I., Moat, H. S., & Preis, T. (2016) Royal Society Open Science, 3(4), 160146.
Quantifying Scenic Areas Using Crowdsourced Data
Seresinhe, C. I., Moat, H. S., & Preis, T. (2018) Environment and Planning B: Urban Analytics and City Science, 45(3), 567-582.
Policy Brief: Using Deep Learning to Quantify the Beauty of Outdoor Places
Chanuki Seresinhe, Suzy Moat and Tobias Preis,
Warwick Business School, University of Warwick
PhD Thesis: From landscapes to cityscapes: Quantifying the connection between scenic beauty and human wellbeing
PROFESSIONAL EXPERIENCE
Head of Data Science, Zoopla & Hometrack
April 2022 - Present
Since April 2022, I have led data science innovation at Zoopla and Hometrack, spearheading projects that boost user engagement and offer critical insights into the UK real estate market.
Zoopla: Elevating User Experience Through Machine Learning
Data Science Foundation: Established the data science function from the ground up, including the development of a sophisticated ML platform. This foundational work has enabled advanced analytics and machine learning capabilities across the organisation.
User Interest Profiling: Utilised machine learning to analyse user behaviour, effectively identifying profiles such as high-intent buyers and sellers. This strategic insight has significantly refined the property search and discovery experience.
Recommendation Systems: Created ML-based recommendations for similar properties and areas, greatly enhancing user satisfaction and engagement.
Generative AI for Area Guides: Innovated with generative AI and human-in-the-loop processes to produce in-depth area guides, enriching user knowledge about different locales and boosting organic site traffic and lead generation.
Hometrack: Refining Automated Valuation Models (AVM)
ML Integration: Strategically integrated machine learning to bolster AVM accuracy, delivering property valuations of unmatched precision to the top mortgage lenders in the UK. This focused application of ML has significantly enhanced valuation reliability, setting a high benchmark within the mortgage industry.
Property Condition Analysis: Pioneered the use of explainable AI in NLP and computer vision to accurately determine property condition, which has traditionally been a task performed through physical inspections.
Advanced AI-Driven Property Insights
Smart Property Tags: Initiated a ground-breaking project to harness the capabilities of computer vision and NLP, developing a system poised to identify key property attributes such as “Period Property,” “Kitchen Island,” and “Sea View” from images and descriptions. Powered by advanced models like BERT for text analysis and convolutional neural networks for image processing, this innovation is set to significantly enrich our property data. Its upcoming full implementation promises to transform the property search and discovery process, enabling more nuanced and user-specific recommendations, and unlocking new insights into property trends and preferences.
Director of Data Science at Culture Trip
August 2020 - May 2022
Leading a team of data scientists, ML Engineers, and Data Engineers to create on a host of ML-driven projects including recommender systems, SEO optimisation and automatic content categorisation at Culture Trip - the travel e-commerce brand with content at its core.
Lead Data Scientist at Popsa
April 2019 - August 2020
Using artificial intelligence to automatically curate people’s unorganised photography content into beautifully designed physical products.
Senior Data Scientist at Channel 4
October 2018 - March 2019
Building Machine Learning/Artificial Intelligence products to help improve the Channel 4 viewer experience.
–
Alan Turing Institute (national institute for data science and artificial intelligence), PhD enrichment progamme & Visiting researcher
September 2016 - Present
Applying deep learning techniques to images of urban locations in order to explore which visual features of the environment might impact the wellbeing of city residents.
Data Science Lab, Warwick Business School, University of Warwick, Doctoral researcher
September 2014 - June 2018
My PhD involved using big online data sets, from such sources as Flickr and Twitter, and deep learning to understand how the aesthetics of the built and natural environment impacts our wellbeing. Teaching activities include mentoring on the FutureLearn MOOC Big Data: Measuring and Predicting Human Behaviour.
–
Digital Design Consultant
2010 - 2014
Advising clients on the creative and strategic use of digital technologies (while I returned to university to study business economics and behavioural and economic science and eventually pursued a PhD).
Micha Weidmann Studio / Design Science Office, Director & Co-Founder
2003 - 2009
A digital design consultancy specialising in identities, websites, and online and offline publications for the arts, culture, and design. Clients included the Design Museum, the Royal Academy, Tate, Dezeen, Fornasetti, Modus Publicity and Zaha Hadid Architects.
I directed the work environment where projects flourished, in turn delivering the end results so valued by our clients. Some of our key achievements included:
Following the redesign of Dezeen, visitors to the leading design and architecture online magazine more than doubled, reaching over 650,000 a month. Universally praised by leading figures in the industry, the redesigned Dezeen made Design Week’s Hot 50 and Time magazine’s Design 100 list of the most influential forces in global design.
Content strategy and design consultation for the Design Museum Shop resulted in a tenfold increase in monthly revenue within six months. In collaboration with search engine optimisation experts Koded, we helped the Design Museum achieve a top-5 position in Google for the coveted keyword “design”.
Recollective, Director & Co-Founder
2001 - 2003
An award-winning online design consultancy with a focus on creating accessible websites backed by high-quality design principles for clients such as NESTA, the Audit Commission and Surrey County Council. My main role included project management and design of digital identities and interactive applications.
Deepend, Designer
2000 - 2001
Design of digital identities and interactive applications on such award-winning projects as the Getty Images Bloodbank (featured in the D&AD annual and nominated for a BIMA) and Conran.com (winner of both a BIMA and a London International Award).
Parsons School of Design, Paris, France, Lecturer
1999 – 2000
Lecturer of web design and development course.
Lumiere.com/Fashion.net, Paris, France, Art Director
1999 – 2000
Design direction and online strategy for online fashion magazines, Fashion Net and Lumiere.
Agency.com, New York, designer
1998 – 1999
Design of digital identities and interactive applications for clients such as Gucci, Armstrong, and Johnson & Johnson.