cv

B.A, M.A. & MEng with Honours

University of Cambridge

Information and Computer Engineering

• Computational Statistics
• Machine Learning
• Deep Learning
• Statistical Signal Analysis
• Data Transmission
• Information Theory and Coding

Data Scientist

The Alan Turing Institute

• Researching AI Security and Safety within the Turing’s Defence and National Security programme, with UK Government stakeholders and academic collaborators
• Demonstrating and presenting my AI research at AIUK, Royal Astronomical Society, Government research workshops and showcases, as well as at academic conferences and stakeholder meetings
• Serving as a technical advisor to shape Ministry of Defence (MoD) guidance on Dependable AI in Defence (JSP 936)
• Contributing to the open-source MITRE ATLAS matrix following collaborative research on ‘Vulnerabilities in Critical Governmental Use of Large Language Model-based Analysis Tools’
• Engaging in cross-organizational collaboration to co-organise the Women in AI Security Workshop 2024 and 2025, bringing together experts across government, academia, and industry
• Winning first place at the UK MoD x Google Cloud Hackathon- the judging panel consisted of Senior HM Government leaders and Google’s technology leadership, including GCHQ’s Chief Data Scientist, Head of DAIC, and Google Cloud’s Vice President of Engineering
• Building leading-edge Reinforcement Learning systems through close collaboration with stakeholders at the Foreign Office to improve detection of possible hostile activity in UK’s Arctic Mission

Research Fellow

Supervised Program for Alignment Research

• Leading applied mechanistic interpretability research under the mentorship of Andy Arditi (former Anthropic AI Safety Fellow), and publishing the research at ICML 2025 R2-FM Workshop
• Winning the first place $2,000 prize for my research fellowship presentation on ‘Adversarial Manipulation of Reasoning Models using Internal Representations’

Research Community Manager

Climate Change AI (CCAI)

• Reviewing academic papers submitted to CCAI’s workshops on ‘Tackling Climate Change with Machine Learning’ at leading international AI/ML conferences, ICML and NeurIPS
• Managing a rapidly growing community platform of +7, 000 members which serves to catalyse impactful work at the intersection of climate change and machine learning by driving interdisciplinary collaboration between technical and subject matter experts

Research Software Engineer (Intern)

Okinawa Institute of Science and Technology Graduate University

• Providing practical embedded hardware and software support for multi-purpose, multi-sensor autonomous systems at the Neural Computation Unit under Professor Kenji Doya
• Developing software (Python, Java, Git) to efficiently and robustly implement the model-based policy search learning framework, ‘Probabilistic Inference for Particle-Based Policy Search’ (Parmas et al., 2018), for an ambitious real-world robot balancing task
• Exercising good coding practices when writing, testing and deploying production code, such as design pattern implementation, Git version control, technical documentation, issue tracking and unit testing

AI Student Researcher

Cambridge University Robotics x Google DeepMind

• Advancing reinforcement learning research in collaboration with Deepmind and the Bio-Inspired Robotics Laboratory, which required exercising my problem-solving skills and ability to learn quickly
• Developing edge AI methods in a simulated environment, adapted from OpenAI gym
• Collaborating with research colleagues and technical experts from Google DeepMind to apply novel AI algorithms, engineering and embedded hardware techniques

Innovation Lead

Impact Through Innovation Cambridge

• Managing an audacious portfolio of high impact projects in an interdisciplinary student team, through a culture of rapid prototyping and relentless problem-solving for innovating under ambiguity

Embedded Systems Intern

City, University of London

• Embodying the laboratory’s cutting-edge research outputs through rapid hardware prototyping, programming, constructing electrical circuits, supervised under Professor Panicos Kyriacou

First place, $2,000 prize

Supervised Program for Alignment Research (SPAR)

Awarded for my research fellowship presentation on 'Adversarial Manipulation of Reasoning Models using Internal Representations' (Yamaguchi et al., 2025)

First place at the UK MoD x Google Cloud Hackathon

UK MoD and Google Cloud

Awarded for my team's project on semantic scanning of cyber security threats. The judging panel consisted of Senior HM Government leaders and Google’s technology leadership, including GCHQ’s Chief Data Scientist, Head of DAIC, and Google Cloud’s Vice President of Engineering

Diamond Jubilee Scholarship

The Institution of Engineering and Technology

Awarded a 4-year merit-based scholarship by the Institution of Engineering and Technology, for the duration of my undergraduate and master's degrees at the University of Cambridge