PolyChord Ltd is a spin-out from Cambridge University astrophysics department. We are a data science company with unique technology which can tackle big challenging complex datasets that defeat other tools. The PolyChord technology was originally invented to extract information from the 10 billion year datasets acquired by the Planck Satellite. It is a unique take on John Skilling's Nested Sampling and has many applications in the real world, including parameter optimisation in complex processes and feature extraction. We have also been using PolyChord to make a new kind of Principled Machine Learning, deploying the core technology to make informed choices about weights and architectures in neural networks. The company is now fast growing and working in areas analysing Life Sciences, protein folding food sciences, POC sensor devices, Additive Manufacturing (3D printing) and clean energy.
Opportunities for permanent roles in a range of business sectors
PolyChord is a fast-growing Cambridge-based data science company (also with a small London office) with genuinely unique cutting-edge tools, a spinout from the Cambridge University Astrophysics department in 2017. We have just been awarded our second Innovate UK grant and seek to employ a data scientist to work either from home, or as things develop (according to personal preferences in this respect) in central Cambridge at The Hauser Centre which is part of IdeaSpace in the midst of Cambridge's thriving new technology sector. Ideal candidates would have a PhD in Physics, Astrophysics or Computer Science and a desire to work alongside some of Britain’s most talented data scientists, with a new technology.
We envisage working initially starting on our first Innovate UK project and also encompassing our second one. In the first Innovate UK project we are working alongside National Research Council of Canada, Queens University Belfast (Professor Seamus Fanning) and Canadian University of Guelph (Professor Lawrence Goodridge), using PolyChord’s abilities to bette rpredict the likelihood of pathogens occurring in foods and food storage. This project has blossomed recently and has led to interesting new areas of investigation which the commercial world highly values. In the second project we are using PolyChord’s abilities to more fully map challenging and complex data landscapes to interrogate infrastructure through sensor data. We would like to move forward to predictive maintenance for which PolyChord is well suited.
We get invited to a lot of Networking events across a wide-ranging span of different applied technologies, including clean and renewable energy, connected and autonomous vehicles, Satellite based Earth Observation; and Life Sciences. Prospective candidates should feel comfortable with (and hopefully enjoy) attending such events, engaging with potential collaborating organisations and helping to explore opportunities - obviously only after a considerable period of initial familiarisation.
- Analyse data in Innovate UK funded FoodScan project using core PolyChord tool working alongside leading data scientist PolyChord Chief Technical Officer Dr Will Handley
- Analyse data in second Innovate UK funded WM5g project using PolyChord tool as above but this time mapping complex data landscapes and producing predictive maintenance outputs
- Work to develop integrations of PolyChord technology - using python
- Collaborate with field testing team to develop analytical prototypes through
- to production
- Assist with creation of web based interface
- Contribute to a series of other interesting and challenging data science in
- different commercial use cases.
- PhD (Physics, astrophysics, Statistics, Mathematics, Computer Science, etc.)
- Possibly 1 - 2 years of experience in quantitative analytics or data modelling
- Some understanding of (and interest in) predictive modelling, machine-learning, possibly some knowledge of clustering and classification techniques, and algorithms, ideally some knowledge of Bayesian model fitting
- Fluency in some programming languages eg (Python, C,C++, Java, SQL)
- Familiarity with Big Data frameworks and knowledge of issues relating to dimensionality and variables in modelling, possibly some knowledge of visualization tools
- A genuine interest and enthusiasm for Bayesian data science and interest in cutting-edge approach to solving problems in mathematics /physics and a desire to bring Machine Learning forward in new and more effective ways.
Salary depending on qualifications and experience