At AstraZeneca we’re dedicated to being a Great Place to Work. Working here means being entrepreneurial, thinking big and working together to make the impossible a reality. We’re focused on the potential of science to address the
unmet needs of patients around the world. We commit to those areas where we think we can really change the course of medicine and bring new ideas to life.

Summary

We are seeking a senior AI Technical Business Partner to lead the industrialization of AI across our R&D organization. This is not just a prototyping role; it is a scaling role. You will bridge the gap between scientific intent and enterprise-grade software execution. You will define the AI-ready data strategy, partner deeply with our data and AI Infrastructure teams, and ensure that our AI solutions transition from isolated pilots to robust, secure, and scalable corporate assets.

Mission

To operationalize intelligence at scale. You will drive the transition of AI from "experimental notebooks" to "enterprise capabilities," ensuring that our data foundation is AI-ready and our technical architecture is secure, compliant, and integrated with the global ecosystem.

Key Responsibilities

1. Technical Discovery & Rapid Prototyping (The "Builder")

• Build Proof-of-Concepts (PoCs): Don't just write a requirements doc; build the MVP. Use Python, LangChain, or low-code orchestration platforms to spin up functional prototypes (e.g., a RAG-based literature review bot or a molecule property predictor) to validate use cases with scientists immediately.

• Feasibility Analysis: Evaluate incoming requests not just for business value, but for technical reality. Assess data readiness, API availability, and model suitability (e.g., "Can Llama-3 handle this toxicology reasoning, or do we need a fine-tuned BioMistral model?").

• Sandbox Management: Manage local AI environments for your therapeutic area, ensuring scientists have secure, compliant access to test new models.

2. Solution Architecture & Agent Orchestration (The "Architect")

• Design Agentic Workflows: configure the specific "tools" and permission sets for AI agents. Define exactly how an autonomous agent interacts with internal APIs (e.g., querying the ELN, accessing chemical inventory) using standards like the Model Context Protocol (MCP).

• Data Engineering Liaison: Write complex SQL queries or Python scripts to assess data quality before a project starts. Translate scientific data needs (e.g., unstructured pathology reports) into structured engineering tickets for the core Data Platform team.

• Integration Strategy: Map the technical integration of AI tools into existing scientific software (e.g., integrating a generative design tool directly into Schrödinger Maestro or Benchling).

3. Governance & Technical Risk (The "Engineer's Conscience")

• Red Teaming & Evaluation: Personally "red team" internal models to identify hallucination risks in scientific outputs. Set up automated evaluation frameworks (LLM-as-a-Judge) to monitor model drift in clinical workflows.

• Security Implementation: Enforce technical guardrails to prevent PII/PHI leakage in prompt injection attacks. Ensure all prototypes comply with GxP and 21 CFR Part 11 standards regarding audit trails and reproducibility.

Qualifications

Technical & Strategic Experience:

• Scaling AI: Proven track record of taking at least 2-3 AI/ML products from concept to enterprise-wide deployment (not just local pilots). Experience with MLOps best practices (MLflow, Kubeflow, model registry).

• Data Strategy: Deep understanding of data engineering for AI. Experience designing "Feature Stores" or "Knowledge Graphs" that feed AI models. Familiarity with scientific data standards (HL7, FHIR, CDISC) is a strong plus.

• Regulatory & HGR Knowledge: Experience working with sensitive healthcare data (PII/PHI) and understanding the complexities of Human Genetic Resources (HGR) regulations (e.g., GDPR, China HGRAC, HIPAA) in a global R&D context.

Technical Stack (Hands-on familiarity required):

• Cloud & Infra: Strong grasp of cloud architectures (AWS/Tencent/Ali) and containerization (Docker/Kubernetes). You don't need to be a DevOps engineer, but you must speak their language.

• AI Platforms: Experience with enterprise AI platforms (e.g., Databricks, Snowflake) and vector database architectures.

• Coding: Proficiency in Python for data analysis and prototyping.

Soft Skills:

• Cross-Functional Leadership: Ability to drive consensus across "triads" of stakeholders: Scientists (Users), Engineers (Builders), and Security/Legal (Guardians).

• Strategic Communication: Ability to explain to the C-suite why "cleaning data" is a capital investment required for "AI success."

Why Join Us?

• Architect the Future: You won't just use tools; you will help define the corporate AI infrastructure for the next decade of drug discovery.

• High-Stakes Data: Work with one of the world's most valuable datasets—proprietary genetic and clinical data—that has the potential to cure diseases.

• Empowered Role: This role sits at the nexus of decision-making, with the mandate to shape technology, process, and culture simultaneously.