Keystream are supporting an NHS organisation to seek an experienced
Digital Research Solution Architect.
The role is £400/ £500 per day Outside IR35
Hyrid role - twice a week on site in London
Start: ASAP
Duration: 12 months
to lead the architecture and design of cutting-edge
AI-enabled research platforms within a highly regulated healthcare and research environment.
This role will play a key part in shaping the future of digital research capabilities, designing scalable and secure solutions across
cloud and on-premise infrastructure, with a strong emphasis on
AI/ML platforms, MLOps, Large Language Models (LLMs), and research-focused data architecture.
The successful candidate will work closely with research, clinical, digital, infrastructure, cyber security, and delivery teams to translate complex research and business requirements into secure, compliant, and scalable technical solutions.
Key Responsibilities
- Lead end-to-end solution architecture and technical design for digital research and AI-enabled platforms.
- Define architecture principles across cloud, on-premise, and hybrid environments, ensuring scalability, resilience, and security-by-design.
- Design and support AI and machine learning architecture, including MLOps delivery frameworks, model lifecycle considerations, governance, and deployment patterns.
- Provide expertise around Large Language Models (LLMs) and Generative AI, including infrastructure considerations, integration approaches, data boundaries, context management, and operational implications.
- Design secure architectures aligned to research governance, data protection, and compliance frameworks, ensuring auditability and responsible AI principles are embedded.
- Collaborate with infrastructure teams on compute, networking, storage, and virtualisation architecture, including GPU-enabled workloads and modern hypervisor platforms.
- Support architecture across both BAU and delivery-focused initiatives, balancing strategic design with practical implementation.
- Work closely with delivery and engineering teams to ensure architectural principles are successfully translated into operational solutions.
- Contribute to architecture standards, technical governance, and roadmap development for digital research capabilities.
Essential Experience & Skills
- Proven experience working as a Solution Architect / Technical Architect / Enterprise Architect within complex enterprise environments.
- Strong experience designing AI-enabled or research-focused technology platforms.
- Demonstrable exposure to Machine Learning Operations (MLOps) including deployment models, validation, governance, and operational processes.
- Experience with Large Language Models (LLMs) / Generative AI, with an understanding of infrastructure and architectural considerations.
- Strong cloud architecture experience (Azure, AWS or hybrid cloud environments).
- Experience designing secure-by-design architectures, including IAM, access controls, compliance, and governance.
- Knowledge of on-premise infrastructure, networking principles, and hybrid delivery models.
- Experience working alongside infrastructure teams supporting virtualisation / hypervisor technologies (e.g., Hyper-V, Proxmox, VMware, GPU-enabled environments).
- Strong stakeholder engagement skills, with the ability to bridge technical and non-technical audiences.
- Experience working within highly regulated environments (NHS, healthcare, life sciences, pharmaceutical, research, or similar).
Highly Desirable
- NHS, research, clinical, or academic health sector experience.
- Experience within research platforms, clinical systems, genomics, imaging, or data-intensive environments.
- Knowledge of responsible AI frameworks, model governance, and AI ethics.
- Familiarity with GPU-enabled compute environments and infrastructure for AI workloads.
- Experience supporting both BAU and project delivery architecture.
If you have the relevant experience for this role - please send your cv for immediate review.
Please note, if you do not hear within a few weeks please note your application was unsuccessful on this occasion.