Gabryel Mason-Williams

Gabryel Mason-Williams

PhD student Computer Science and Research Software Associate

Queen Mary Univeristy of London

Rosalind Franklin Institute

Biography

I am doing a PhD in Computer Science at Queen Mary University of London, supervised by Mark Sandler and Primoz Skraba, researching the Theory of Deep learning in the aim to make Deep Learning more explainable and interpretable. I am also in a contract role with the Rosalind Franklin Institute, working on a novel lossless neural compression model for scientific data.

Download my resumé.

Interests
  • Theory of Deep Learning
  • Uncertainty Qunatification in Machine Learning
  • Neural Networks Pruning
  • Explainable AI
  • Scientific Applications of Artificial Intelligence
Education
  • PhD in Computer Science, Present

    Queen Mary University of London

  • MSc in Artificial Intelligence, 2023

    Queen Mary University of London

  • AI and Machine Learning in Healthcare Summer School, 2022

    Cambridge Centre for AI in Medicine

  • BSc in Computer Science, 2021

    University of Plymouth

Accomplish­ments

Google 2023 Academic Scholarship
€7000 for demonstrating passion for technology, academic excellence, and exceptional leadership.
Principle’s PhD Scholarship
Covers tuition fees, and provides annual tax-free maintenance for 3 years.
Scicomm Hackathon
This was a hackathon organised through CERN to build, develop and prototype projects in the field of scientific communication. I was the leader of the project as selected by the group. Our project was to create and design a tool to explain how a project’s goals and achievements evolve over time. This was an interdisciplinary team made of communicators, designers, programmers and scientists.
Diamond Light Source - Year In Industry Certificate
An award for completing my project “Processing Big Data- High performance object stores for tomography data processing.
Deans List
An award for achieving more than 70% in an academic year

Skills

Research Software Enginnering

Developing novel research software utilising the Scientific Python Stack (Numpy, Scipy, Matplotlib) and the Python Machine Learning Stack (Jax, Pytorch and Tensorflow).

Teaching and Mentoring

Taught students from year 8 to undergrad in computer science and maths

Interdisciplinary Research

A large section of my research is at the intersection of multiple disciplines

Experience

 
 
 
 
 
Research Software Engineer
Sep 2023 – Present Oxfordshire
At the Rosalind Franklin Institute (RFI), I am presently involved in the research and creation of novel neural losses compression algorithms for large volumetric scientific data.
 
 
 
 
 
Research Software Associate
Jul 2022 – Sep 2023 Oxfordshire
At the Rosalind Franklin Institute (RFI), I am presently involved in two research projects. DisTRaX, which I lead, is the productionisation and improvement of previous research and Kompressor, a machine learning project to create a novel neural losses compression algorithm for extensive volumetric scientific data.
 
 
 
 
 
Junior Research Software Engineer
Jul 2021 – Jul 2022 Oxfordshire
In this role I conducted research in High-performance computing (HPC) for big data processing. Within this research, I lead the development and maintenance of two significant projects, DosNa and DisTRaC. I was also involved in writing a successful grant bid for £10,000 in research funding for “Using Novel Cluster Technology to improve performance of Cryo EM analysis” with the University of Bristol. This was successful and lead to a reduction in processing times of 4.37%.
 
 
 
 
 
AI & I Research Project- Automating Lab Book Data Collection With Edge Machine Learning
Sep 2020 – Jun 2021 Oxfordshire
The dissertation/research project aimed to develop an automated way to collect lab book information and automatically record it in a central database. This object was achieved using classical computer vision techniques to recognise when to take a photo of the lab book and machine learning to generate a synthetic handwritten keyword dataset for training and testing and to extract the handwritten keywords from the lab book. The data captured was then stored in an s3 data store, and words extracted were attached in the metadata of the uploaded images. This dissertation was awarded 81.68% and is currently under further development at the Rosalind Franklin Institute.
 
 
 
 
 
Maths Circle Mentor
Oct 2021 – Jul 2022 Exeter
I mentored students in years 8 and 9 in mathematics to encourage the development of mathematical thinking outside of school context- the aim is to improve their ability to be adventurous, articulate and accepting by providing them challenging questions to become better mathematicians.
 
 
 
 
 
Scientific Computing Research Placement
Jun 2019 – Sep 2020 Oxfordshire
I worked on a 12-month research project entitled “High Performance Object Stores for Intermediate Data processing” this involved using technologies such as High Performance Compute Clusters, Univa Grid Engine, Savu, Ceph and relevant deployment tools as well as programming technologies such as MPI, C, python and bash. During this time, I deployed and maintained a semi-production Ceph cluster using ansible, created a way to visualise objects with an object store which formed part of a display at SC19, made GRAM a kernel module to represent RAM as a block device for Ceph and DisTRaC an application that allows Ceph to run on to on a High Performance Computing cluster using RAM. I also presented a poster at Ceph Day CERN as well as giving well-attended talks on the work within Diamond and the Ceph Science Group.

Recent Posts

Popular Topics