Data Scientist
Send a job offer directly to this candidate
Former Data Scientist Intern with Masters in AI and Computer Science
Data Scientist at Frontier Economics
◦ Key Collaborations Across 3 Practice areas: Energy, Competition, Public Policy:
∗ Led a customer segmentation project utilising Unsupervised Hierarchical clustering algorithm on energy sector survey data, achieving a refined understanding of customer behaviour. Executed data pre-processing with one-hot encoding and standardisation, followed by ’complete’ linkage and ’Hamming’ distance in the clustering algorithm. Delivered 10 distinct clusters with a 25% improvement in silhouette score, ensuring cluster cohesion.
Also reduced within-cluster variance by 30%, confirming compactness.
∗ Initiated and optimised the Geospatial analysis process for 10 major M&A transactions at a Private Equity firm, enhancing acquisition evaluations efficiency.
Responsibilities included data cleansing, creating specialised directories, and implementing a Local Overlaps Model for market share analysis. Played a key role in liaising with legal teams and the CMA, ensuring transactional compliance.
Python and ArcGIS to automate catchment and drivetime analysis, improving geographical and service accessibility assessments of veterinary practices. Achieved a 40% reduction in evaluation time and enhanced market share calculation accuracy by 25%, significantly contributing to the firm’s strategic growth. My work effectively integrated technical analysis with legal and regulatory compliance.
∗ Developed and implemented an advanced data analysis and visualisation system using R and Leaflet, achieving a 30% increase in the accuracy of London’s public transport network accessibility evaluations. My role entailed processing the TfL’s Numbat dataset to devise a novel station crowdedness index, based on the number of platforms as a proxy for capacity. Specifically, I calculated the ratio of boarders and alighters to the number of station platforms during AM peak times, identifying Victoria Station as the busiest with 100% crowding.
This benchmark allowed for a linear rescaling of other stations’ crowdedness relative to Victoria, integrating this data into the PTAL assessment to reflect varying levels of station accessibility and connectivity. The result was a significant enhancement in the PTAL measure, now inclusive of a refined crowdedness measure, leading to more accurate and informed transport planning, policymaking, and resource allocation across London’s transport network.
Masters in AI University of St Andrews:
Courses: Machine learning, Signal Processing (Computer Vision), Artificial intelligence (Algorithms and Data Structures), Information visualisation, Digital Heritage(Augmented Reality).
Masters in CS University of Newcastle
Courses: Software Development Techniques, Databases, Web Technologies, Computer Networks, Advanced Software development
