PolyChord Ltd is a data-science company, which spun-out from the University of Cambridge's Astrophysics group in order to commercialise a unique next-generation data-science technology.
As a consultancy company, we leverage over 30 years of expertise to effectively address complex blue-sky and commercial data-analytics challenges. PolyChord applies cutting-edge techniques transferred from the ever-advancing world of academia, to deliver world-class data-analytics solutions. We make ourselves commercially available through fixed consultancy fees, or in joint collaboration with organisations, sharing the commercial outcomes.
Since 2017, we have worked with a diverse range of clients to deliver cutting-edge tailored solutions that provide the end user with a competitive advantage. Our extensive expertise ensures data-analysis challenges are met with speed, precision, and cost effectiveness - driving meaningful results no matter the complexity of your needs. PolyChord delivers explicable results, which other competing approaches struggle to deliver, allowing our clients to apply our tools with confidence.
Learn moreWe work closely with every client from the outset, combining their field-specific expertise with our world-class data-science knowledge to deliver the optimal solution to the problem. We have worked with the likes of the Defence Science and Technology Laboratory (DSTL), Transport for London, West Midlands Metro, and the UK Atomic Energy Authority on a wide variety of problems. These projects have resulted in a number of widely applicable tools, that deliver reliable and effective results that deliver game-changing impact on the ground.
PolyChord have specialised tools that model out noise and unwanted signal variations to reveal the true signal buried in your data. With this cleaned data we apply advanced statistical inference and machine learning techniques to provide evaluation of the current condition of infrastructure and probabilistic predictions of its remaining operational lifetime. This tool has been used by the likes of TfL and National Highways to allow them to monitor the health of bridge structures using low-cost accelerometers that can be widely deployed.
Many functional surfaces that industry needs to optimise exhibit multiple hills and valleys, and/or long thin ridges. The PolyChord sampler is uniquely capable of fully optimising such complicated functions, even in the presence of complex constraints on the permitted sampling space, and for 100s of input variables. For such surfaces, standard samplers will generally only find a few of the regions of interest and will miss important information relating to the problem at hand, whereas PolyChord will deliver a full-coverage sampling.
PolyChord’s cutting-edge data science applies multiple methods to robustly identify genuine anomalies and outliers to identify where there is a potential issue that needs to be inspected and fixed. Our tool provides a prioritised list so that resources can be sent to where they are needed most. This tool has been used by companies such as West Midlands Metro and LNER to automatically detect and locate issues on track, vehicles, pantographs and OLE.
Is your simulation fast, but not fast enough for involved analysis, such as investigation of the full input-parameter space, or to make robust predictions of future system behaviour? Our advanced sampler delivers a full-coverage AI surrogate that delivers robust and complete answers to all your complex questions in seconds, rather than in days or weeks. This tool returns almost real-time simulation output and has been applied by AVL Advanced Simulation Technologies.
Working with the Defence Science and Technology Laboratory (DSTL), we have developed MIDAS - a cutting-edge Bayesian methodology for deciding what kind of data should be next be collected in order to deliver optimal results. We have already showcased MIDAS beating human opponents in a simple game of capture the flag. This tool can be used to plan iterative data-collection campaigns, and for the design of optimal sensor networks, as well as robust and easily-interpretable information using their data.
PolyChord is transforming the landscape of small-molecule analysis with Molytics, our cutting-edge technology that reveals the lowest energy state of molecules and small-molecule complexes. Unique to our full-coverage sampling of the energy function, is the provision of information about the dynamical movement of the molecules and complexes, and the ability to probe the molecular structure and dynamics at any temperature. This tool is driving sustainable innovation across critical industries such as mining.
Chief Science Officer
Mike Hobson is a world-renowned physicist and statistician with over 30 years experience. Mike is a key inventor of PolyChord’s cutting-edge Bayesian technology. He has authored 500+ publications on machine learning and Bayesian statistics (including the acclaimed book ‘Bayesian Methods in Cosmology’).
Email: [email protected]
Chief Technology Officer
Will Handley is an expert in Bayesian data science and the main inventor of the PolyChord technology. Will has overseen many successful projects including work with the Ministry of Defence, National Highways and TfL. He has authored 150+ papers relating to Bayesian statistics and machine learning since 2014.
Chief Executive Officer
Mike Handley brings 40+ years experience in business and commerce to PolyChord Ltd. Mike successfully founded and led PolyChord through dynamic growth where it is now employing 11 people. While Mike has been CEO, PolyChord has raised over £2m in UK-government funding.
Email: [email protected]
Lead Data Scientist
Catherine Watkinson has over 14 years experience working in applied data science. Catherine has worked at world-class institutes like Imperial College London (for both PhD & several postdoc positions) and University College London. She has successfully managed >7 major innovation projects and is a Certified Scrum Master.
Email: [email protected]
Our technical team is headed up by two University of Cambridge Professors, Professor Mike Hobson and Associate Professor Will Handley, who are both world-renowned experts in machine learning and statistics. Together, they ensure that our solutions are at the forefront of AI and statistical innovation
Our data-science team is led by Dr. Catherine Watkinson who has over 14 years experience in applied machine-learning and data science. All of our data scientists hold postdoctoral qualifications, many from world-renowned institutions such as Imperial College London, University College London, and the University of Cambridge. Their expertise spans a range of disciplines, including engineering, computational science, and astrophysics, with each having specialised in data science within their respective fields.
Click headshots for more info.
If you have a data problem that you think PolyChord could help you with, contact us using the form below or email [email protected] to set up an initial exploratory meeting.
Contact us today and begin turning complex data into explainable and easily-interpretted information that enables robust decision making. We have various models of working with clients, from simple fee-paid work to output-sharing collaborations.