Data Scientist (consultant)
NHS England / Crown Prosecution Service
Skills
C++, Flask, Machine, TensorFlow, Scikit-Learn, Weka, Spark MLlib, Docker, Scipy, Numpy, Azure ML, AWS, Google Cloud Platform, GPUs, GPU clusters, Windows, Mac, Linux, Unix, Android, GPU, machine learning, recommender system, deep learning, LSTM, word2vec, CVs, Google, Computer Vision, neural network, convolutional neural networks, dialogue systems, Matlab, Perl, Mathematica
Project history
06/2020
-
06/2020
* Predictive models for staff attrition and staff training outcomes for NHS employees and for
criminal cases.
criminal cases.
04/2019
-
02/2020
Data Scientist
Tesco plc
* Designed, developed, trained and deployed two large scale machine learning solutions.
* Vehicle route planning for 4000 stores around the UK.
* Order prediction for 15 million home shopping customers based on purchase history.
* Vehicle route planning for 4000 stores around the UK.
* Order prediction for 15 million home shopping customers based on purchase history.
04/2018
-
03/2019
Data Scientist
CV Library
* Principal data scientist at job board with 12 million jobseekers.
* Developed recommender system for sending job alerts to candidates, with 9% conversion rate.
* Trained deep learning models (CNN, RNN, LSTM, word2vec, seq2seq) to analyse candidates' CVs
and job descriptions, using Google GPU instances.
* Deployed model to analyse candidates' CVs in real time when they register, giving better
candidate experience and a lower bounce rate on the registration form.
* Developed recommender system for sending job alerts to candidates, with 9% conversion rate.
* Trained deep learning models (CNN, RNN, LSTM, word2vec, seq2seq) to analyse candidates' CVs
and job descriptions, using Google GPU instances.
* Deployed model to analyse candidates' CVs in real time when they register, giving better
candidate experience and a lower bounce rate on the registration form.
04/2017
-
03/2018
Computer Vision Scientist
Veridium
* Managed a project to develop neural network face recogniser.
* Trained convolutional neural networks with TensorFlow, able to classify images such as
fingerprints or pharmaceutical pill bottles.
* Designed and patented cryptographic measures to protect biometric data.
* Trained convolutional neural networks with TensorFlow, able to classify images such as
fingerprints or pharmaceutical pill bottles.
* Designed and patented cryptographic measures to protect biometric data.
05/2011
-
10/2015
Knowledge Engineer / Pre-sales consultant
Artificial Solutions UK; Pattern Science AG
* Developed human-like dialogue systems for use on mobile and in consumer electronics.
May 2011-Oct 2015
Solution architect - Pattern Science AG, near Frankfurt am Main, Germany
* Machine learning for monitoring market sentiment
May 2011-Oct 2015
Solution architect - Pattern Science AG, near Frankfurt am Main, Germany
* Machine learning for monitoring market sentiment
08/2009
-
03/2011
Research assistant / Teaching assistant - Psychology Dept.
Birkbeck College
* Analysed and processed MRI data sets using Matlab and Perl.
07/2006
-
08/2006
Research assistant
European Synchrotron Radiation Facility
* Simulated X-ray scattering on atoms using Mathematica.
Jul-Aug 2006
Jul-Aug 2006
07/2005
-
08/2005
Research assistant
Center for Applied Mathematics and Theoretical Physics
* Publication: Testing adiabatic invariance in separatrix crossing (Robnik & Wood, 2006).
Jul-Aug 2005
Jul-Aug 2005
Local Availability
Open to travel worldwide
