Contractor specialising in a range of healthcare data, technology, and professional services

Advisory Role (Part-Time) • Provided advisory services in data science model development and training, focusing on non-pharmaceutical intervention coding,• Supported a finance company in implementing and fine-tuning large language models for analyzing free text data and extracting product insights, using Retrieval-Augmented Generation (RAG) to improve the accuracy and relevance of results.

Advisory Role (Part-Time)Serving as a fellow providing advisory support to startups at Zinc, a venture capital firm fostering new companies aiming to solve significant societal problems through innovative solutions.

(Part-Time)Serving as a fellow in the NHS Clinical Entrepreneur Programme, receiving training and mentorship to develop innovative health-care solutions within the NHS framework

Lead Data Architect for the Regulatory Management System/Regulatory Connect Project, I was responsible for the comprehensive data strategy and leadership across several critical areas managing a team of up-to 16 Data Modellers and Data Analysts. My duties covered the data architecture for the Product Licensing Application, Process Licensing Application, and Medical Device Registration(including Software as a Medical Device), Import Notification System, Inspections, Customer Account Management, Master and Reference Data (including MedDra and ATC classifications, substances), Site/Facilities, Finance, etc. My role also extended to overseeing the development of both Conceptual and Logical data models to ensure seamless operation and compliance across the system. I was responsible for maintaining IDMP and HL7 FHIR standards to enhance interoperability between different data systems. This involved coordinating with a team dedicated to upholding these standards, ensuring that all processes and data handling were in line with regulatory requirements.Contract was via CloudKubed

Lead Data Architect at the Medicines and Healthcare Products Regulatory Agency (MHRA)Leading requirements gathering, design, solution architecture, and data architecture of healthcare enterprise data solutions

Advisory Role (Part-Time)Advisory role supporting the development of digital solutions for infection prevention and control, leveraging wearable technology (RTLS) to track and manage pathogen spread in healthcare settings employing Graph Neural Networks and Social Network Analysis.

Co-led development an AI tool for automating review moderation on NHS.UK, leveraging BERT, GPT-Neo, Hugging Face, and spaCy.• Developed models in azure ML and deployed models using Flask, Azure Kubernetes Service (AKS), Azure Container Apps and developed in Azure DevOps integrated with the .NET solution• Implemented MLOps strategy for scalable, robust deployment optimized for seamless, low-latency performance

• Led development of a data architecture strategy for a Japanese pharmaceutical firm, enabling Population Health Management insights for market access.• Led automation of internal processes using Python (Pytesseract, Selenium), automating PDF extraction via OCR and Regex, and written to excell files.• Architected a BI solution for a Japanese medtech firm using Azure (Logic Apps, Functions) and Power BI, developing a sales dashboard and reporting solution.• Enhanced interoperability between supplier and ePMA systems by leading an NLP project using Hugging Face, spaCy, and BioBERT to match data across non-compatible systems.

• Managed and led the technical sales account strategy for Databricks’s largest UK client, a multinational pharmaceutical company, providing support with Solution Architectures, Data Architectures, and the design of federated data models to improve interoperability.• Aided bioinformatics projects, including GWAS analysis with Spark for parallel processing, and implemented NLP for named entity recognition of side effects in phase 4 clinical trials.• Applied AI techniques in manufacturing and commercial sectors, implementing computer vision for quality control and developing forecasting models for Fast-Moving Consumer Goods (FMCG) to optimize stock replenishment for a consumer goods client.• Supported Databricks’ interoperability templates, including HL7 V2/3, FHIR, and OMOP, to streamline data integration and support large-scale healthcare initiatives.

• Led development of a demand and capacity planning tool for a governmental healthcare body, optimizing patient pathways and resource allocation through Discrete Event Simulation.• Developed and deployed NLP machine learning models on GCP for post-stroke speech rehabilitation, providing real-time speech analysis and feedback to enhance recovery.• Engineered an R Shiny application to support hospital planning for colonoscopy units, enabling the input of operational variables to forecast the required number of units based on projected demand, enhancing efficiency and resource management in healthcare facilities.

Led the design and deployment of AI and Data Science solutions across multiple projects at Hitachi Vantara and Hitachi Europe. Managed and executed key projects including:– Led data science for Hospital Digital Control Centre initiative that provided predictive insights into patient flow, significantly improving operational response times and patient care delivery. Developed in Azure ML, deployed using AKS and ACI via Azure DevOps.– A predictive maintenance solution for a Danish mining company using AWS Sagemaker, which utilized unsupervised time-series models to predict equipment deterioration.– Development of a patented solution architecture for patient pathway variation analytics, enabling healthcare staff to quickly identify and address inefficiencies impacting hospital KPIs.– Process mining of cancer pathways to uncover insights on why certain patients were not managed within the NHS 62-day target, providing actionable recommendations for improving adherence to this key performance indicator (KPI).– Strategy and implementation of a population health management project for a major NHS Trust, evaluating data and deploying models to optimize resource allocation and patient care strategies.– Conducted research and development in areas such as autonomous vehicles using computer vision and localization of 3D point clouds from Lidar data, and innovative biometric solutions leveraging computer vision for fingerprint vein recognition in a fintech application.

Led the design and deployment of AI and Data Science solutions across multiple projects at Hitachi Vantara and Hitachi Europe. Managed and executed key projects including:– Led data science for Hospital Digital Control Centre initiative that provided predictive insights into patient flow, significantly improving operational response times and patient care delivery. Developed in Azure ML, deployed using AKS and ACI via Azure DevOps.– A predictive maintenance solution for a Danish mining company using AWS Sagemaker, which utilized unsupervised time-series models to predict equipment deterioration.– Development of a patented solution architecture for patient pathway variation analytics, enabling healthcare staff to quickly identify and address inefficiencies impacting hospital KPIs.– Process mining of cancer pathways to uncover insights on why certain patients were not managed within the NHS 62-day target, providing actionable recommendations for improving adherence to this key performance indicator (KPI).– Strategy and implementation of a population health management project for a major NHS Trust, evaluating data and deploying models to optimize resource allocation and patient care strategies.– Conducted research and development in areas such as autonomous vehicles using computer vision and localization of 3D point clouds from Lidar data, and innovative biometric solutions leveraging computer vision for fingerprint vein recognition in a fintech application.

Working on dataset with millions of patients, I used Unsupervised Machine Learning Algorithms to cluster patients with diagnosed Type-2 Diabetes into three groups in a precision medicine approach. The clusters were then presented for long term outcomes using survival analysis and visualised in a patient-patient 3-D network