Job Description

Lead the optimization of our NLP pipelines for clinical text processing, balancing performance, cost, and accuracy in a healthcare setting Integrating NLP pipelines with structured databases like OMOP and other SQL databases Pioneer the future of clinical AI by developing and implementing novel LLM-based solutions for clinical applications, from automated documentation to decision support Design and build robust, scalable architectures for deploying LLM systems across our hospital network, ensuring they meet NHS security and privacy standards Develop strategies for effective knowledge retrieval from diverse clinical sources, including EMRs, clinical guidelines, and medical literature Collaborate with clinical teams to identify and solve complex healthcare challenges using state-of-the-art language models Guide the evaluation and implementation of both commercial (e.g., OpenAI, Anthropic) and open-source LLMs (e.g., Llama, Mistral) for specific clinical use cases Leading the strategic development and implementation of NLP pipelines to extract structured information from unstructured clinical notes, radiology reports, pathology reports, and other text-based medical documentation. Managing and coordinating a team of NLP engineers and data scientists to develop standardized approaches for clinical text processing across the trust. Designing and implementing advanced NLP solutions using state-of-the-art language models while ensuring patient privacy and data security. Leading the integration of extracted insights from clinical text into existing data workflows and clinical decision support systems. Regularly meetings with the Head of Data Science and Clinical Director for AI, Data and Innovation Lead to report on team progress, negotiate with, review and prioritise the scope of the applications/infrastructure in accordance with the Trusts wider strategic plans. Supervising and mentoring junior NLP engineers and data scientists, fostering knowledge sharing and technical excellence within the team. Developing and maintaining documentation standards for NLP models, including model cards, data specifications, and validation metrics. Working closely with clinical teams to validate NLP outputs and ensure clinical accuracy and relevance. Establishing best practices for text preprocessing, annotation, and quality assurance in healthcare NLP applications.

This role sits within the AI, Data & Digital Innovation function of the Chief Medical Officer's Office (CMO). It is made up of data science, technology and digital transformation experts, operating as part of the wider King Health partner's (KHP) group. This role sits with the NLP & Analytics team, based in GSTT but working closely with teams at KCH and KCL.



The teams work on some externally funded programmes - that develop and implement Data platforms and AI with the aim to achieve better outcomes for patients' diagnosis and treatment, improve clinical and operational workflows and drive innovative healthcare research. This includes the London Artificial Intelligence Centre for Value-Based Healthcare - a consortium of NHS, academic and industry partners led by Guy's & St Thomas', King's College London and KCH in partnership with 10 NHS Trusts and 4 universities.



Mentor and grow a team of NLP engineers while fostering a culture of innovation and clinical impact



Leading the development of evaluation frameworks to assess NLP model performance in clinical settings. Presenting complex NLP solutions and their impact on clinical care to various stakeholders, from technical teams to hospital leadership. Working within a team, act as an expert individual contributor for architecting and developing software solutions that are currently in an early research stage. For example - Machine Learning Decision Support tools for clinical decision support and population health, software for Clinical Trial Recruitment, and software for Agentic Clinical Decision Support. This may include de novo development, or refactoring prototype code for production deployment, following software engineering best practices. Work with KHP Stakeholders in refactoring front-end of software used by Trust data and analytics teams (e.g. AI Centre, Trust EDW/SDE, CogStack) Working within a team, act as an expert individual contributor to the software codebase of this federated analytics and machine learning platform, following software engineering best practices. Working with both clinical data science, clinical safety, and ML engineering teams, as well as clinical end-users, to ensure that solutions are impactful and safe.

Communications:

Communicate with engineering and infrastructure staff on a regular basis to agree design and implementation plans.



Collaborate regularly with clinical teams and infrastructure staff to align NLP system design with healthcare needs and technical constraints.



Lead engagement with clinical departments across GSTT to build consensus on NLP applications and drive adoption of AI-powered solutions.



Maintain productive relationships with clinical teams, data scientists, and infrastructure teams to ensure successful deployment of NLP products - including LLM systems, RAG pipelines, and autonomous clinical agents.



Provide detailed performance metrics and analysis of NLP systems, including model accuracy, latency, and clinical impact metrics to the Head of AI/ML on a weekly basis.



Manage relationships with LLM providers, clinical NLP tool vendors, and academic partners.



Department to review NLP system performance and validate clinical accuracy



Chair technical workshops and design reviews to align NLP development across multiple clinical specialties and use cases at GSTT.



Present complex NLP architectures and their clinical impact in an accessible format to various stakeholders, from clinicians to hospital leadership.



Communicate model results, limitations, and potential biases to clinical teams; guide implementation decisions; explain technical concepts clearly to non-technical audiences.



Analysis and Judgement



To be a Subject Matter Expert (SME) for the designated area to be able to support peers and the post holder's direct reports.



Ensure that NLP systems adhere to clinical safety standards, NHS data protection requirements, and emerging AI governance frameworks.



Make judgements on how to design and implement core configuration within the technology stack.



Trouble shoot, determine root cause and provide system solutions for assigned modules.



Utilise expert knowledge and experience of workflows and processes to develop pathways and new processes to support the configuration and implementation of the designated application which will impact on working practices across the Trust.



Analyse and document highly complex clinical and administrative



workflows, information, data collection and/or reporting requirements.



Analyse and review test activity, progress and variances to define



recommendations to maintain schedules and raise any risks or issues



highlighted through testing to the appropriate level whilst ensuring thorough



tracking of data.

Planning:

Work with the designated Programme/Project Manager/other members of the teams using PRINCE2 and/or Agile principles, to ensure your team meets the set deadlines to enable go-live.



Work with the Director of Data Science and other KHP stakeholders to formulate and document the core setup plans.



Work with clinical stakeholders and other end-users around London to design product lifecycles that end in impactful and measurable outcomes.



Review and amend allocated elements of the plans and provide updates to the Head of Data Science.



Management responsibility



Responsible for leadership and development of work-packages and plans associated with role;



Assign specific designs of the module set up in accordance with their complexity to appropriately skilled application analysts;



Technical management and leadership of a geographically wide team in the design and build of NLP platforms and related products. This will include contributing to recruitment, performance management, managing grievance and disciplinary procedures, mentoring and coaching where appropriate;



Ensure that the work of the team conforms to the design integrity for the designated application.