Lead ML Engineer, driving research to improve RFID localisation performance.• Designed & implemented full Data Science pipeline & stack for collaborative RFID modelling.• Developed 2-tier regression models boosting speed & accuracy, contributing to a significant increase in RFID localisation efficiency.

Empowering businesses to unlock hidden potential through strategic data insights. Leverage expertise in Machine Learning, NLP, and Data Analysis to deliver custom solutions that tackle complex challenges and drive tangible results. Collaborate closely with clients to understand their unique needs and translate them into actionable strategies, fostering a partnership for impactful data-driven success.

Under my capacity as a Senior Data Scientist Contractor at HR DataHub in London since August 2023, I spearheaded the development of a Docker-based web UI that leveraged Active Learning to significantly improve data labelilng efficiency and address dataset imbalance. This innovative solution featured a dynamic Confusion Matrix, PR Curves, F1 score, Precision, Recall, and offered deep statistical insights to enhance decision-making processes. Additionally, I played a crucial role in updating the MongoDB repository by identifying and integrating critical benefit fields. This enhancement streamlined backend services and enabled seamless frontend user access and search functionalities, thereby optimising the overall user experience and operational efficiency.My accomplishments within this role are following:• Spearheaded the development and integration of NLP and Data Science solutions, analysing benefits within a MongoDB repository of 24 million job descriptions.• Advanced identification of benefits using BERT (LLM) models, achieving superior performance over traditional tree-based and linear models with Tf-idf and Doc2Vec, evidenced by a 0.9 Precision and Recall metrics.

As a key developer and Senior Data Scientist at IG Group (2020-2023), I played a primary role in impactful projects across Financial Risk Management, Client Care, and Trading, leveraging my expertise in Machine Learning, NLP, and data science to drive tangible results.Key Achievements:Financial Risk Management:• Played a primary role in the technical development of a Hedging Simulator utilising Luigi, to empower Quants to test strategies efficiently. This tool incorporated client exposure prediction capabilities, and resulting in very significant cost savings.• Designed and implemented a dynamic markets correlation matrix, providing valuable insights for optimising budget allocation and risk management strategies.Client Experience and Support Optimisation:• Revolutionised client complaint categorisation with an NLP model (BERT, LLM), reducing resolution times and support costs.• Developed a system using Random Forests to identify high-value clients, enabling proactive engagement and loyalty building, bolstering revenue generation from VIPs.Content Impact Analysis:• Applied Deep Learning techniques, including Long Short-Term Memory (LSTM), to develop an attribution model for content touchpoints, optimising client engagement strategies and enhancing client journey experiences.Talent Development and Knowledge Sharing:• Led a successful research papers reading club, fostering continuous learning and knowledge sharing.• Actively mentored colleagues and co-organised Kaggle challenges, promoting a culture of growth within the team.

As a Data Scientist specialising in Threat Identification at Digital Shadows, I played a key role in developing and implementing AI-powered solutions to detect and mitigate online threats. Leveraging Machine Learning, NLP, and Computer Vision, I contributed to several impactful projects, including:• Automated Threat Detection: Developed and deployed systems using big data analytics to automatically detect various threats (typosquatting, domain squatting, vulnerabilities) across the open and dark web, protecting clients from digital risk, data loss, and online brand security.• Advanced Logo Detection: Engineered highly accurate logo detection capabilities through Convolutional Neural Networks (CNN), significantly enhancing brand protection measures.• Data Loss Prevention: Utilised NLP to detect shellcode and classified-marked documents, preventing data breaches and leaks.• Cybersecurity Investigations: Analysed sextortion campaigns and other threats, contributing to improved cybersecurity measures.• Contributed to the development of a tech method that had a provisional patent filed for.Beyond technical expertise, I actively fostered a culture of learning by launching an internal tech training program, recognised as the "Best initiative ever at Digital Shadows".

As a Data Science Research Engineer, I played a key role in developing innovative NLP and Deep Learning solutions to revolutionise ad targeting. I contributed significantly to projects that enhanced both efficiency and personalisation, leading to improved client reach, engagement, and ad performance.Key Contributions:• Conceptualised and designed a self-adapting NLP algorithm, capable of classifying billions of user searches for targeted ads and adapting to dynamic keyword taxonomies, even with rapidly changing categories. This eliminated the need for ongoing data collection and model retraining, saving resources and ensuring scalability.• Implemented advanced linguistic analysis techniques (PMI collocations, part-of-speech tagging, sentiment analysis) to refine search relevance and ad targeting, ensuring users saw personalised, relevant ads.• Contributed to the development of innovative user profiling methodologies using word embeddings and CNNs, categorising users for highly targeted, personalised ad experiences.• Conceptualised, developed, and presented an image analysis approach on ad creatives using Convolutional Neural Networks (CNNs) with transfer learning to evaluate its impact on Click-Through Rates (CTR) at an internal hackathon. This solution achieved a top 4/30 ranking, demonstrating the potential for my proposed approach to significantly improve ad performance.

Octave / Matlab, Unix scripting. Tutoring large groups of students in lab workshops

C++, Unix scripting. Tutoring large groups of students in lab workshops

Software Engineering module. In charge of small groups of 2-4 students, helping Cambridge undergraduate students to prepare for their exams. Responsible for preparing the structure and materials for each session, assignments, marking, and providing individual feedback to their Director of Studies

CORBA, Unix scripting. Tutoring large groups of students in lab workshops

Java, Android, Unix scripting. Tutoring large groups of students in lab workshops

As a Systems Integrator intern at Intel Corporation, I actively sought opportunities to contribute to the team's efficiency and streamline processes. Utilising my skills in C# and SQL Server, I participated in projects that led to significant time savings and workflow improvements.Key Achievements:* Identified a repetitive manual task requiring 2 hours of daily team effort. I contributed to the development of an automated solution using C# and SQL Server, reducing daily workload by 25% for our team, freeing up valuable time for more strategic tasks.* Collaborated with another intern from the marketing department to address their manual Excel sheet processing challenges. I contributed to the creation of a solution comprised of a VBA macro, an SQL database, and a batch script, significantly reducing their workload and ensuring data accuracy and consistency.* Proactively presented project proposals to relevant stakeholders, effectively communicating the benefits and gaining support for implementation. This experience helped me develop essential communication and collaboration skills.