Data Engineer

InCrowd (rebranded to Cortex), Apr '24 – Present

• Lead site reliability for 260+ Prefect flows supporting all ETL pipelines, maintaining a 99.7% success rate across 5,000+ daily jobs.
• Collaborated on the migration of Arsenal FC's data warehouse from PostgreSQL to Snowflake, reducing cloud costs and enabling self-service analytics.
• Engineered and deployed ETL pipelines across 9 fan-data platforms serving 15 clients, including Crystal Palace, RFL and EuroLeague (Python, Prefect, Docker, Kubernetes, EC2).
• Led Norwich FC's ticketing provider migration from Advanced to SeatGeek, spanning ETL and data-warehouse modelling.
• Designed and productionised third-party data integrations (Salesforce, Flowcode, DotDigital, OneSignal, VenueMaster, Experian, Adobe Magento), syncing 100M+ records monthly.
• Contributed to data-warehouse modelling to unify digital engagement, transactional and gamification data, supporting the transition to a SaaS product (dbt, PostgreSQL, Snowflake).
• Co-engineered real-time audit-log emission for event-driven flows using Apache Kafka.
• Co-implemented gRPC-based Prefect signal logging to enable real-time monitoring and alerting in Datadog.
• Delivered hundreds of client features across 30+ organisations, including automated SFTP exports, seasonal reporting flows and BI logic updates.
• Built GDPR-compliant data flows, including automated Right to Removal and Subject Access Request pipelines.
• Led Snowplow data-archival strategy and client offboarding processes (S3, Glacier, Redshift).

Data Scientist

InCrowd, Jul '21 – Mar '24

Promoted from Junior Data Scientist (including part-time while studying).

• Produced AWS cloud-cost forecasting from CloudFront usage for 30 clients across 100s of microservices, achieving a 30% cloud-cost reduction and accurate billing.
• Developed and deployed a client-facing co-occurrence analytics tool using association rules (Lift) to uncover relationships between user demographics and purchasing behaviour (Streamlit).
• Designed engagement-scoring models to quantify fandom levels, enabling audience segmentation for marketing and gamification strategies.
• Led the Data Science team in establishing experimentation standards, CI/CD practices and model-lifecycle tooling (MLflow, PyCaret, SageMaker Feature Store).
• Built a Text2SQL self-service tool powered by LLMs, enabling non-technical teams to generate analytics independently and reducing manual analyst workload by 90%.
• Developed NLP pipelines for automated article tagging and categorisation using topic modelling and NER to drive retail and content promotion.
• Built internal data-quality tools measuring completeness, nullity and cardinality (Great Expectations).
• Built classification models to predict season-ticket and subscription renewals for Crystal Palace (LTV, churn modelling).
• Developed a campaign-optimisation pipeline for audience targeting across retail, ticketing and memberships, increasing click-through rates by 10%.

Industrial Placement — Data Engineering & Analytics

InCrowd, Aug '19 – Aug '20

• Delivered monthly and seasonal client-facing reports and rapid-turnaround ad hoc analysis with Account and Product Managers.
• Modelled raw Snowplow and GA4 event data into analytics-ready warehouse tables using dbt.
• Engineered cross-database pipelines leveraging a central data lake (Dremio, PostgreSQL, Redshift, MySQL).
• Built Tableau dashboards visualising KPIs across demographics, digital engagement and purchasing behaviour.

Junior Research Associate — Natural Language Processing

TAG Lab / CASM, Jun '19 – Aug '19

• Used SOTA transformer LMs (BERT) to establish text-processing pipelines powering client applications.
• Implemented large-scale scraping and semantic-analysis pipelines to extract political arguments from Twitter and Reddit.
• Constructed word-embedding networks and clustering analysis using Gephi to map ideological spectrums for Parlia.com.

MSc Data Science — Part Time, Distinction

University of Sussex, Sep '21 – Jun '25

Rigorous postgraduate training in the mathematical foundations of ML and AI — probability, statistical inference, linear algebra, calculus and optimisation. Dissertation: techniques to overcome token-length constraints in transformer-based language models for long-document classification.

BSc (Hons) Computer Science and Artificial Intelligence — 2.1, with Industrial Placement

University of Sussex, Sep '17 – Jun '21

BCS-accredited degree combining software engineering and machine learning with research in AI (Natural Language Engineering, Computer Vision, Neural Networks). Dissertation: ALT (Article Library Toolkit).

Machine Learning Engineering for Production (MLOps)

Coursera, Jun '24

User Interviews, Design & Wireframe — Bank Group Project

University of Sussex, Nov '20

First prize — Machine Learning Hackathon

HackSussex, Nov '19

Workshop — Solidity blockchain smart contracts

HackSussex, Jul '19

Backend Development — C# Cluedo SWE Group Project

University of Sussex, May '19

IT Support Assistant

OMM, Jun '15 – Sep '17