Senior AI Engineer/Scientist

Senior AI Engineer/Scientist

Contract (Outside IR35)

Remote

Until end of Dec 2026 (Initial, likely extensions)

Competitive rate (DOE)

About the Role

Exalto Consulting are supporting a leading digital health and technology organisation in the search for an experienced Senior AI Engineer/Scientist.

This is a senior, hands-on role focused on designing and building production-grade agentic AI systems for the structured authoring of clinical study documents. You will develop end-to-end, multi-agent systems that coordinate activities such as planning, drafting, evidence ingestion, validation, and compliance checks across complex clinical content workflows.

The role sits at the intersection of applied AI, NLP, software engineering, and cloud architecture. We are looking for someone who can go beyond proof-of-concept LLM applications and engineer robust AI capabilities that improve reasoning quality, traceability, domain alignment, and the reuse of accurate content across clinical development.

Key Responsibilities

Agentic AI & Multi-Agent Systems

• Design, build, and optimise agentic and multi-agent AI systems for clinical document authoring.
• Develop workflows that coordinate planning, drafting, evidence ingestion, validation, and compliance-checking agents.
• Improve agent reasoning quality, robustness, traceability, and alignment with clinical-domain requirements.
• Design approaches for grounding generated content in source evidence and producing accurate, reusable content blocks.
• Evaluate and improve LLM and agent behaviour using appropriate testing, evaluation, guardrail, and observability techniques.

NLP, Information Processing & Document Intelligence

• Build capabilities to extract, structure, retrieve, transform, and generate information from complex documents and datasets.
• Apply both classical and modern NLP techniques where appropriate, including information extraction, retrieval, embeddings, and LLM-based approaches.
• Develop reliable approaches to evidence ingestion, provenance, and traceability across AI-generated clinical content.
• Work with structured and unstructured information to support accurate and reusable authoring workflows.

Production AI Engineering

• Engineer AI/ML capabilities for production use rather than prototype-only or API-wrapper solutions.
• Build maintainable Python-based services and components with strong software engineering practices.
• Deploy and operate AI/ML/LLM systems with appropriate testing, monitoring, reliability, and scalability.
• Collaborate with engineering, data, product, and domain teams to integrate AI capabilities into end-to-end production workflows.

Cloud, MLOps & Lifecycle

• Design and implement cloud-native AI solutions, with significant emphasis on AWS.
• Contribute to serverless and event-driven architectures supporting scalable AI workflows.
• Apply MLOps and DevOps practices across deployment, monitoring, versioning, and continuous improvement.
• Support model and data lifecycle management, including reproducibility, lineage, evaluation, and controlled change.

What You'll Bring

• Strong hands-on experience building GenAI and agentic AI solutions, including genuine multi-agent or agent-orchestrated workflows.
• Strong Python and software engineering capability, with experience building robust production systems.
• Experience in NLP, document intelligence, or information processing, particularly extracting, structuring, retrieving, and generating information from complex documents.
• Proven experience deploying and operating AI, ML, or LLM systems in production environments.
• Significant cloud architecture experience, preferably AWS, with serverless and event-driven architectures particularly relevant.
• Experience with MLOps, DevOps, model/data lifecycle management, LLM evaluation, guardrails, observability, or RAG would strengthen the profile.
• Experience in life sciences, clinical development, healthcare, or another regulated industry would be advantageous.
• Ability to work effectively across technical and domain teams and translate complex requirements into reliable AI capabilities.

Interested?

If this sounds relevant, we'd be happy to share more context and discuss whether it could be a good fit.