Background
I specialise data science and AI, and bring strategic thinking and clear communication to technical problems.
Read more Read less
I started as a quantitative researcher at Cambridge University Press & Assessment in 2022, moving into a data scientist role via an MSc apprenticeship in data science. In 2026 I joined the BBC as a data scientist.
Before that I studied social science at the London School of Economics and the University of Bristol. Alongside statistical training and a lot of writing, this taught me to think in systems: to look at human behaviour and structural problems critically and at scale. It is a habit of mind I still use daily.
I also spent time teaching, in the UK and internationally, which meant explaining difficult ideas to people encountering them for the first time. I developed my listening and understanding of others further as a Samaritans listening volunteer for eleven months in 2024.
Data Scientist
BBC
I am a full-stack data scientist, working across the whole lifecycle from research and development through to deploying and monitoring of models in production.
Read more Read less
This includes:
- Writing Python to process data and to train and evaluate ML models and LLMs.
- Building on AWS: sourcing data, training models, integrating model APIs, and running inference for live events.
- Working inside a cross-functional product team alongside the software engineers and testers who maintain the platforms these models are deployed on.
- Working with business stakeholders to make sure what we build meets what they actually need.
Data Scientist
Cambridge University Press & Assessment
I worked as a data scientist on LLM work, research, and data engineering.
Read more Read less
This included:
- Piloting an AI bid assistant that used GPT models on Microsoft Azure to help bid managers prepare funding applications.
- Building a Python pipeline that automated materials costing and procurement for around 20 million books a year.
- Publishing research on automarking, AI ethics, and AI in the classroom.
- Monitoring and resolving failures in the data processing jobs behind exam operations.
- Developing automarkers for short text responses, training random forest models and fine-tuning BERT models.
- Using LIME and SHAP to assess the explainability and construct relevance of automarkers.
Researcher
Cambridge University Press & Assessment
Lead researcher on a quantitative scoping review of the literature on LLMs and automarking.
Read more Read less
This included:
- Synthesising quantitative evaluations of LLMs such as BERT and GPT-4, comparing performance across precision, recall, F1, and macro-F1.
- Presenting research on detecting AI-written coursework to Ofqual, the exams regulator for England.
- Producing metadata for the Research Division's reports and migrating them to Apollo, the University of Cambridge repository, so the Division's research is indexed and discoverable.
Research Assistant
Cambridge University Press & Assessment
While studying for an MSc Degree Apprenticeship in data science, I worked on a range of quantitative research projects.
Read more Read less
This included:
- Automating a 120-page competitor analysis of grade boundaries in R Markdown, so the report could be regenerated annually rather than rebuilt by hand.
- Developing new metrics to measure consistency in examiners' marking annotations.
- Supporting data processing and grade boundary setting during OCR's high-stakes exam series.
Researcher
Dixon International Group
Researcher and executive assistant to the managing director of a fire safety manufacturer, covering data analysis, research, and project management.
Read more Read less
This included:
- Querying the company's SAP system to run exploratory analyses, with processing and visualisation in R.
- Project-managing the specification, procurement and installation of perimeter fencing to improve site security.
- Preparing research and briefing material to support the managing director's decision-making.
Teacher
Teach First
Secondary school teacher on the Teach First leadership programme, teaching five classes of GCSE and pre-GCSE students.
Read more Read less
This included:
- Planning, delivering and marking more than fifteen lessons a week.
- Breaking down difficult concepts, Earth Overshoot Day for instance, into terms teenagers could use, drawing on frameworks such as ADEPT (analogy, diagram, example, plain English, technical definition).
- Managing classroom behaviour through relationship-building, consistent application of the school's behaviour policy, and Teach Like a Champion techniques.
Teacher
My English School, Naples
English teacher for professionals in Naples, Italy.
Read more Read less
This included:
- Delivering four to five one hour lessons per day for up to ten adult learners.
- Teaching across the full range of levels, from absolute beginner to advanced.
- Providing tailored feedback and one-to-one tuition.
- Gaining extensive one-to-one mentoring on my teaching and communication.
Teacher
American Forum, Buenos Aires
Freelance English teacher and private tutor for professionals in Buenos Aires, Argentina.
Read more Read less
This included:
- Building a freelance teaching practice from scratch, finding and retaining my own clients.
- Teaching several lessons a day, mostly business English, including to politicians and finance executives.
- Preparing students for the IELTS and CAE qualifications.
MSc Data Analytics
Anglia Ruskin University · Distinction
Postgraduate apprenticeship in data science and machine learning, taken part-time alongside full-time work.
Read more Read less
Dissertation: Comparing neural networks and tree-based models in predicting sales conversion yield.
Year 1 modules:
- Exploratory Data Analysis
- Data Engineering and Big Data
- Machine Learning Techniques
Year 2 modules:
- Advanced Time Series Analysis
- Deep Learning and Applications
MSc Policy Research
University of Bristol · Merit
Rigorous quantitative social research training, applying statistical methods to real-world policy questions.
Read more Read less
Dissertation: Using linear regression analysis to predict risk of secondary teachers leaving the profession
Modules:
- Philosophy and Research Design in the Social Sciences
- Introduction to Qualitative Research Methods in the Social Sciences
- Introduction to Quantitative Research Methods in the Social Sciences
- Further Quantitative Methods
- Further Qualitative Methods
- Researching Poverty, Inequality and Social Exclusion
BSc Sociology
London School of Economics and Political Science · 2:1
Broad grounding in social research design, combining quantitative and qualitative methods at one of the world's leading social science institutions.
Read more Read less
Dissertation: Qualitative research through semi-structured interviews on why interviewees in Dagenham, London voted for Leave in the EU Referendum.
Year 1 modules:
- Social Theory
- Power, Inequality, and Difference: Contemporary Themes in Sociology
- Social Psychology
- Statistics in Society
Year 2 modules:
- Key Issues in Sociological Analysis
- Researching London: Methods for Social Research
- Political Geography of Development and the South
- Political Sociology
Year 3 modules:
- Civil Liberties and Human Rights
- Sociology of Race and Ethnicity
- Law and Violence
- Reactionary Radicalism: Populism and Authoritarianism in the 21st Century
Co-founder
Cambridge AI Reading Group · Cambridge
Co-founded a reading group in Cambridge covering foundational and new papers in AI.
Read more Read less
This included:
- Organising and facilitating conversations between AI practicioners in Cambridge (from Microsoft Research, Arm, and GitHub) once a month.
- Reading foundational papers in AI including on perceptrons, backpropagaton, CNNs, embeddings, and the transformer.
- Reading more recent papers on mechanistic interpretability and AI alignment.
- Reading books on AI by Nick Bostrom and Eliezer Yudowsky.
Founder
Cambridge Hiking Club
Founded and nurtured an active hiking community in Cambridge.
Read more Read less
This included:
- Personally organising 20 days of walking both locally and during trips to the Lake District, Peak District, and other locatons.
- Setting up the informal consistution and administration for the group, including rules, norms, the bank account, and the Executive Committee.
- Creating an effective and welcoming culture in the club, encouraging socials, participation, and active engagement.
- Growing the club to 100 active paying members on WhatsApp and 1,000 passive members on Meetup.com.
- Handing over the leadership of the club to the Executive Committee.
Listening Volunteer
Samaritans
Listening volunteer for the Samaritans mental health and suicide prevention helpline.
Read more Read less
This included:
- Undertaking three months of training from experts in effective communication, and benefiting from extensive 1-2-1 mentoring.
- Doing a weekly four hour shift for eight months where I emotionally supported callers. This included night shifts every four to six weeks.