Feb 2023 – Sep 2025 · 2 yrs 7 mos · Greater London, England, United Kingdom · Hybrid

• Spearheaded the development of AI-powered connected car propositions and modern data architecture, securing C-suite alignment and positioning The AA as a thought leader in mobility innovation:
• Defined and executed AA-X’s data strategy, aligning data assets with corporate objectives
• Cross collaborated with product and engineering teams to deliver data services seamlessly, ensuring smooth integration and execution.
• Built and scaled a multidisciplinary AI/ML function (engineers, scientists, analysts), driving innovation and delivery excellence
• Led a team of 5 data scientists to deploy scalable machine learning models, doubling conversion rates, enhancing customer experience, and boosting operational efficiency
• Successfully launched AI powered solutions for both B2C and B2B markets, growing adoption of proactive vehicle health care services
• Established a modern data architecture enabling real-time analytics, advanced insights, and AI integration
• Represented The AA at industry conferences and forums, positioning the company as a thought leader in connected vehicle AI
• Secured C-suite buy in for strategic initiatives, ensuring measurable business impact and alignment with growth strategy

Jul 2020 – Feb 2023 · 2 yrs 7 mos · Greater London, England, United Kingdom

• A role within the retail industry covering a wide range of projects.
• Technical lead of the recommendations team
• Oversight and influence of company-wide data science projects
• High-level stakeholder management
• Methodically adopted scrum and agile way of working
• Developed a product similarity model shown to outperform the 3rd party model with an uplift of 68% via internal testing. Customer-facing A/B testing showed an uplift in conversion (3.92%) and CTR (142.86%)
• Development and deployment of a brand recommendation model. A/B test showed a conversion increase of 1%
• Collaboration with data engineering and platforms teams for the implementation of MLOps + best practices
• Creation and development of the data science target architecture used for data and model monitoring
• Mentoring and line management duties

Jan 2020 – Jul 2020 · 6 mos · London

• A role within the dating industry with projects directed at identifying user preferences and improving user safety
• Improvement of user safety via detection of hate symbols and weapons in images
• Identification of user preferences in both objects and facial features
• Model productionisation experimentation on AWS with ECS and elastic inference

Jul 2018 – Jan 2020 · 1 yr 6 mos · London Area, United Kingdom

• Working within the real estate industry focusing on the development of Zoopla's computer vision capabilities and platform.
• Lead the computer vision team to the "most creative hack" award at Zoopla's internal hackathon
• Development of object detection models for listing enrichment
• Trained CNNs for image classification and image fingerprinting. Fingerprinting used to match similar images and therefore identify duplicate listings as well as related images (Fastai, Pytorch)
• Productionisation of the models through containerisation and deployment to ECS (Docker)
• Developed a computer vision platform to characterise listing images and insert them into a cache.
• Several models are deployed in the platform and are all used in the characterisation pipeline (ECS, Lambda, EC2, Redis, Python, Flask, Docker)
• Improved performance of the platform through asynchronous programming and parallel computing. Additionally, to aid performance further, the models scale horizontally according to demand (ECS, Lambda, EC2, Redis, Python, Flask, Docker)

Aug 2016 – Jul 2018 · 1 yr 11 mos · London Area, United Kingdom

• A role within the market research domain, focussing on full stack development of data science tools and ETLs. Some of my key projects/responsibilities were:
• Team projects to fully develop data science tools (CausalImpact, Docker, Flask, Kubernetes, Luigi,
• Python, Redshift, RabbitMQ, React)
• Developed a model to predict whether a sales transaction will be won or lost. This tool is used
• by the finance team to forecast revenue. (Docker, Flask, Kubernetes, Luigi, MongoDB, Python,
• RabbitMQ, S3, Scikit learn, Spark)
• Created a dashboard for internal use, standardising and automating the data extraction procedure.
• (Angular, D3, Docker, Flask, Kubernetes, Luigi, Python, Redshift, S3)
• Researched A/B testing and how it can be applicable to the business, followed by advising other
• teams on the experiment setup (R)
• Increase awareness of the data science team within the company by developing a
• blog for internal use, showcasing the capabilities of the team (Angular, Docker, Flask, Kubernetes,
• MongoDB, Python, S3)

Nov 2014 – Aug 2016 · 1 yr 9 mos · London

• A diverse data science role within the recruitment industry. The role requires a broad spectrum of skills including algorithm development, web development and machine learning. The majority of my time was used to address the needs of the company, some of which are as follows:
• Creation of machine learning models for salary prediction, industry prediction and classification of
• free text followed by visualisation of the results (Scikit-Learn, Python, Angular, D3, JavaScript)
• Management of the current data science databases (mongoDB)
• Construction of pipelines to extract, clean and upload data (Python, mongoDB)
• Creation of taxonomy trees (Python, D3, R) to aid in customer search and used internally for audience building. In addition, algorithms were developed for tree traversal (Python, R)
• Refactored large parts of the Python code base improving performance of the code
• Currently in early stages of implementing a named entity recognition model to extract information from free text (NLTK)
• Full stack development of data science tools (Python, R, PHP, D3, JavaScript)

Sep 2010 – Sep 2014 · 4 yrs

• Created and designed a series of computational routines/algorithms to address and solve problems in biology. Research was focused on the following areas:
• Statistical analysis of biological datasets (R), to identify when a change in protein shape becomes significant
• Designing computational work flows for data analysis (Python). Generated a set of protein models with a hybrid approach and selected models which best fit shape information. Followed by further refinement based on distances between specific protein points.
• Implementing algorithms for shape calculations (Python and C)
• Development of scientific software (TEMPy and Amphitrite)
• Implementing algorithms for high performance computing (CUDA)
• Demonstrated on a number of courses, helping students understand key topics in the following areas:
• Statistics with R
• Structural Bioinformatics
• Bioinformatics

Feb 2010 – Aug 2010 · 6 mos

• The placement at UCSD involved developing:
• Normalised MySQL database
• Web Interface to access, update, add and remove entries in the database
• Python classes to carry out the backend work
• Functionality to search the database with the BLAST software

Jun 2008 – Feb 2010 · 1 yr 8 mos · London, Portsmouth, York