Principal AI Engineer — AI Solutions (Enabling Functions)
Job Description
Principal AI Engineer — AI Solutions (Enabling Functions)
Introduction to role:
Are you the hands-on AI builder who turns sophisticated governance into code and ships production systems that stand up to regulators, auditors, and real users? Can you move from idea to working proof-of-concept in days, then mature it into a resilient, governed application that powers how a global business hires, contracts, spends, and stays compliant?
Based in Barcelona, you will be the technical authority within a high-impact squad, partnering closely with engineering leadership, enterprise technology, data engineering, legal, security, and functional teams across HR, Finance, Procurement, Legal, Audit, Compliance, and Business Development. Your work will deliver auditable, compliant AI from the first commit—helping colleagues operate faster and safer, and ultimately accelerating how we bring life-changing medicines to patients.
In this builder-first role, you design multi-agent architectures, build RAG pipelines over real enterprise corpora, route models across regions to respect sovereignty requirements, and ship governed applications into live environments. You will set engineering standards by example while remaining the most prolific builder on the team.
Accountabilities:
Technical Design and Solution Architecture: Lead the end-to-end technical design for AI solutions across enabling functions—selecting patterns, defining component boundaries, and making build-versus-integrate decisions aligned to the squad’s architectural direction.
High-Consequence Decisions: Make and document model strategy choices (foundation models, fine-tuning, multi-provider orchestration, RAG), integration patterns, and tooling with clear rationale against product requirements and compliance constraints.
Platform and Reuse: Identify and build shared components, reusable agent patterns, and governance instrumentation that speed delivery across the portfolio—championing platform thinking over one-off projects.
Data Sovereignty in Code: Translate sovereignty policy into working infrastructure—implementing model routing, data residency, and hosting decisions that respect jurisdictional boundaries and cross-border transfer requirements.
Hands-On Multi-Agent Delivery: Design and ship multi-agent LLM systems across providers, write the orchestration code, implement sovereignty-aware routing, and ensure safe outcomes with human-gated control where needed.
Production RAG Pipelines: Build governed RAG pipelines with hallucination guards before any user-facing output and immutable audit logging at every stage—from ingestion through to response.
Rapid Prototyping with Production Intent: Deliver POCs in days with clear progression gates to pilot and production, using real data and defined success criteria.
Governance-as-Code: Implement classification/tiering logic, model registration, approval workflows, monitoring pipelines, decommissioning controls, human-in-the-loop oversight, deterministic routing layers, model/data cards, and full audit trails as first-class engineering components.
Multi-Jurisdictional Compliance: Engineer systems that meet data protection laws, AI-specific regulations (e.g., EU risk classifications), and sector-specific operational resilience expectations—coding for divergence to remain defensible under multiple regimes.
Operational Resilience: Bake resilience into architecture with circuit breakers, provider fallbacks, graceful degradation, fail-safe defaults, tested failover paths, and chaos/resilience testing to defined tolerances.
Delivery and MLOps: Own delivery from design to production, scaling, monitoring, and lifecycle management; establish CI/CD, automated testing, performance monitoring, incident response automation, and capacity management.
Data Engineering and Controls: Build pipelines, feature stores, and data products for quality, governance, and reuse; implement data quality, lineage, and controls for financially material data, legally privileged documents, employee-sensitive information, and supplier-confidential data.
Model Quality and Assurance: Implement drift and bias monitoring, lead AI red-teaming, embed fairness and explainability pipelines, manage technical risks, and produce assurance reporting for internal and external oversight.
Scaling Through Reuse and Partnership: Adapt solutions across functions (e.g., from contract intelligence in Legal to supplier risk in Procurement), contribute patterns to enterprise platforms, and integrate external partners at a technical level to meet production standards.
Stakeholder Partnership and Adoption: Serve as a credible hands-on partner to HR, Finance, Procurement, Legal, Audit, and Compliance—running technical discovery, prototyping quickly, iterating on UX, and measuring adoption and impact through KPIs/KRIs.
Enterprise Integration and Standards: Represent delivery in architecture and AI leadership forums, contribute to enterprise AI governance and standards, share patterns, and advocate for the requirements of regulated, process-critical functions.
Essential Skills/Experience:
Built AI applications inside a large, regulated enterprise.
Demonstrated experience delivering AI solutions across multiple enabling functions (HR, Finance, Procurement, Legal, Audit, Compliance).
Significant experience implementing governance controls and embedding operational resilience into your own code and architectures.
Ability to move from idea to working proof-of-concept in days.
Depth in AI governance, operational resilience, and regulatory compliance as a core professional discipline—not adjacent knowledge.
Technical authority who designs multi-agent architectures, writes production code, builds RAG pipelines, and ships governed AI applications into live enterprise environments.
Experience implementing multi-jurisdictional data sovereignty, regulatory divergence, and cross-border AI governance requirements—translating sovereignty policy into working infrastructure code.
Treats regulatory requirements and governance as architecture decisions made at the keyboard, not as checklists applied after the fact.
Proficiency across model strategy (foundation models, fine-tuning, multi-provider orchestration, RAG) and sovereignty-aware model routing.
Leads through mentorship, code quality, and engineering standards while remaining a strong individual technical contributor.
Desirable Skills/Experience:
Track record building shared libraries, reusable agent frameworks, governance middleware, and common instrumentation that accelerate delivery.
Mastery of AI-assisted development to compress delivery timelines while maintaining deliberate human control over architecture, routing, and guardrails.
Hands-on MLOps experience including CI/CD pipelines, automated testing, observability, incident response automation, and capacity management.
Experience implementing fairness testing, explainability pipelines, AI red-teaming, and assurance reporting suitable for auditors and regulators.
Resilience engineering expertise including circuit breakers, provider fallbacks, graceful degradation, impact tolerances, and chaos testing.
Experience implementing classification/tiering logic, model registration systems, approval workflow automation, monitoring pipelines, and lifecycle decommissioning controls.
Proven partnership across enterprise technology, data engineering, legal, information security, and external technology providers to deliver production-grade solutions.
Experience instrumenting adoption and value through KPI/KRI design and UX-driven iteration in production.
Why AstraZeneca:
Join a technology community where engineers, scientists, and compliance experts share the same whiteboard to unlock the potential of data and AI for patients. We are scaling a bold digital strategy with real investment, modern platforms, and the freedom to experiment—hackathons, rapid prototyping, and production-grade engineering live side by side. Here, ambition is matched by support: we value kindness alongside drive, and we do things the right way, building governed, resilient systems that can be trusted at scale. Your craft will shape how a global company operates today and how quickly life-changing medicines reach people tomorrow.
If you are ready to build governed, resilient, multi-jurisdictional AI that moves from idea to enterprise impact at speed, join us in Barcelona to lead with your keyboard and deliver what matters!
Date Posted
18-Jun-2026Closing Date
17-Jun-2026AstraZeneca embraces diversity and equality of opportunity. We are committed to building an inclusive and diverse team representing all backgrounds, with as wide a range of perspectives as possible, and harnessing industry-leading skills. We believe that the more inclusive we are, the better our work will be. We welcome and consider applications to join our team from all qualified candidates, regardless of their characteristics. We comply with all applicable laws and regulations on non-discrimination in employment (and recruitment), as well as work authorization and employment eligibility verification requirements.