Senior Applied AI & ML Engineer

It currently takes over 10 years and $1.3B to develop a drug. More than 70% of that investment goes into clinical trials, yet only ~10% of candidates make it from Phase I to approval.
Evinova - a new health-tech business within the AstraZeneca Group—is here to change the math. We use advanced algorithms and GenAI to aim high: boosting clinical trial success by 20%, cutting development time by 3 years
, and halving study costs.

As Senior Applied AI&ML Engineer, you will prototype and build the systems that make those targets real - blending scientific knowledge, ML and agentic systems to drive transparent and actionable recommendations. You’ll integrate multi‑source data - historical trials and scientific datasets into production systems to improve the process of designing clinical trials and increase their probability of success.

If you're motivated by meaningful problems and comfortable working outside your existing experience, you'll thrive here. We're looking for generalists with scientific backgrounds and some software, data science and ML skills — people who are genuinely curious and think scientifically, committed to continuous learning, and eager to rethink how clinical trials are designed. We value people who build with depth and intention, not just wrap LLM API calls.

• Ph.D. or equivalent professional experience in a relevant scientific or quantitative field with industrial experience (Bioinformatics, Mathematics, Computer Science, Machine Learning, Statistics, or similar)
• Previous industry experience building applied ML/AI systems that have shipped as part of a product and driven measurable business impact.
• Domain knowledge - familiarity with drug development, clinical trial design, or real-world data (EHR, claims, prescriptions)

• Experience of classical ML and NLP
• Hands-on work with generative AI – including prompt engineering, context engineering and multiagent systems and working with managed endpoints (OpenAI, Anthropic, AWS Bedrock) and open-weight models (Hugging Face ecosystem)
• Knowledge of agentic design patterns - planning, memory, tool use/function calling, RAG - with practical experience in evaluation and guardrails appropriate for regulated environments
• Experience working with scientific datasets/ literature.

• Strong Python development skills with production sensibilities (testing, observability, documentation)
• Experience with containers, APIs, and async services (Docker, FastAPI) and CI/CD pipelines (Git Hub Actions)
• Awareness of architectural patterns in deploying applied ML/AI systems in cloud (AWS).

• Ability to translate complex technical work into clear narratives for both technical and non-technical stakeholders
• Experience sharing knowledge with peers and contributing to engineering and data science standards and best practices.

• RAG pipelines at depth - experience building secure, compliant ingestion and retrieval systems with provenance tracking, including web automation, parsing, and document processing
• Agent frameworks - hands-on experience with multi-agent orchestration tools (e.g., Google ADK, Strands Agents, Lang Graph, CrewAI, or equivalents)
• AI-augmented development - effective use of agentic coding assistants (Copilot, Cursor, Claude Code) to accelerate delivery
• Open source or community contributions - published packages, conference talks, or internal framework development
• Startup-pace experience - comfort with ambiguity, rapid iteration, and wearing multiple hats

Location: St Pancras London (3 days per week onsite / 60% overall)

Salary: Competitive + Excellent Benefits!

Evinova draws on AstraZeneca’s deep experience developing novel therapeutics, informed by insights from thousands of patients and clinical researchers. Together, we can accelerate the delivery of life-changing medicines, improve the design and delivery of clinical trials for better patient experiences and outcomes, and think more holistically about patient care before, during, and after treatment.

We know that regulators, healthcare professionals, and care teams at clinical…