duncan@dsanalytics.co.uk
Machine Learning | MMM | Customer Retention | Bayesian Modelling | SimulationLondon-based data science consultant and director of DS Analytics, which I set up in April 2016.
15+ years’ experience building statistical models and analytics tools for decision makers.
Working across a wide range of organisations including IKEA, BBC, Jo Malone London, The Telegraph, McDonald’s, Post Office, M&S, MediaCom, Twitter, Google, Public Health Wales, The Economist, ITV
Expert in building statistical and machine learning models and packages in R and python.
Extensive experience providing R and python training and coaching data science students.
Emphasis on solving business problems and clear communication.
Current areas of interest - Bayesian methods in marketing
Frequent speaker at conferences (next: LondonR in April 2024)
“Duncan smoothly integrated Bayesian statistics, simplifying our understanding of LTV and turning a complex issue into actionable insights.
His commitment to collaboration was evident, fostering a dynamic exchange of ideas. This, coupled with his technical acumen, ensured a streamlined project progression”
PUBLICATIONS
Rapid Point-of-Care Genotyping to Avoid Aminoglycoside-Induced Ototoxicity in Neonatal Intensive Care
Assessing the impact of introducing point-of-care genotyping on normal care delivery.
Published in: JAMA Pediatrics
DEVELOPING COVID-19 TESTING STRATEGIES WITH MANCHESTER UNIVERSITY NHS FOUNDATION TRUST
During the height of the COVID-19 pandemic, we collaborated with medics and mathematicians in the UK and the US to develop and assess various 'group testing’ strategies for screening health care workers.
We developed software in R for simulating the outcome of alternate testing strategies.
Published in: The Journal of Molecular Diagnostics
UTILIZING POINT-OF-CARE DIAGNOSTICS TO MINIMIZE NOSOCOMIAL INFECTION IN THE 2019 NOVEL CORONAVIRUS (SARS-COV-2) PANDEMIC
We collaborated with medics and academics at Manchester University NHS Foundation Trust to assess and visualise the implications of testing different cohorts for COVID-19 using R and Bayes Theorem.
Published in: QJM Journal of Medicine
“Duncan is a first-rate data science expert. An unusual mix of business understanding and deep statistical knowledge. I’m always impressed by his ambitious and original solutions to difficult problems. Highly recommended.”
“It’s been a pleasure working with Duncan and DS Analytics. He grasped the problem quickly and provided truly innovative solutions, far exceeding the brief. Can’t recommend highly enough.”
“DS Analytics were immensely helpful getting our product team up and running with R and data science. Would definitely recommend them.”
EDUCATION
London School of Economics (2011-2012)
MSc Economics Merit; 83% econometrics
Cardiff University (2006 - 2009)
BSc Economics 1st Class Honours
EXPERIENCE
DS Analytics May 2016 - Present (MD & Data Scientist)
Founded data analytics consultancy, specialising in machine learning, forecasting and business simulation modelling. Clients include the Telegraph, BBC, MediaCom, Privatam, Jo Malone, essence and Beano.
AlphaSights May 2015 - May 2016 (Data Scientist)
Providing insight and analytical tools to support one of the fastest growing companies in Europe. Bayesian prospect scoring models, dynamic map visualisation, text mining, new market Monte Carlo simulation, cost forecasting. R & python.
Mindshare UK Sep '12 - May '15 (Consultant)
Data analyst / business consultant working with clients such as M&S, McDonald's, Twitter, Post Office and Volvo. Developing innovative analytical methods to solve client problems. Winning new business (inc. largest in company's history). Building R packages.
Institute of Economic Affairs July 2012 - August 2012 (Intern Researcher)
Research and blogging for renowned thinktank.
Transactis Aug 2010 - Dec 2011 (Analyst)
Building propensity scoring models for direct marketing clients.
duncan@dsanalytics.co.uk