Director, AI/ML

AstraZeneca plc; India; Hybrid; INR 6,000,000 - 9,000,000; Full time

Job summary

AstraZeneca plc in Bangalore seeks a Director of AI/ML to set the technical bar for AI systems shaping therapies. You will be the technical authority across conventional ML to Generative and agentic AI, producing measurable outcomes for patients and the business.

Lead through influence, defining architecture, standards and strategy with hands-on expertise. You will partner with Commercial, Medical and Technology teams to transform diverse data into rigorous AI.

Qualifications

• BS in computer science, mathematics, statistics, engineering, physics or related field.
• Extensive experience (12+ years) building and deploying data science and ML solutions with recent 2+ years on LLM/agentic apps in production.
• Proven track record of setting technical direction and leading AI delivery through influence.

Responsibilities

• Define and drive technical vision, roadmap and architecture for traditional ML, Generative AI and agentic AI.
• Provide hands-on technical leadership to project teams; review architectures and code.
• Design multi-agent systems and LLM applications with tooling, memory, guardrails and human-in-the-loop patterns.

Skills

Generative AI/LLMs

Agentic AI

Evaluation & observability

ML foundations

Programming & tools

Cloud & MLOps

Education

BS in CS / math / engineering

Master’s/PhD in ML/CS (desirable)

Tools

Airflow

Spark

MLflow

Docker

Job description

Job Title: DIRECTOR, AI/ML

GCL: F

Location: Bangalore

Introduction to role:

Are you ready to set the technical bar for AI systems that shape how life-changing therapies reach patients? In this senior individual-contributor role based in Bangalore, you will be the technical authority guiding advanced AI across our commercial landscape-from traditional ML to Generative and agentic AI-turning breakthroughs into measurable outcomes for patients and the business. You will lead through influence, not headcount, defining architecture, standards and strategy while pairing that with hands‑on expertise. Partnering with colleagues across Commercial, Medical and Technology, you will help transform diverse data-patient, physician, health systems, sales and scientific content-into rigorous, scalable AI that drives better decisions at pace. How would you design resilient AI systems that are trusted, governed and production-ready from day one?

Accountabilities:

• Technical Vision and Architecture: Define and drive the technical vision, roadmap and architecture for traditional ML, Generative AI and agentic AI capabilities, aligned to enterprise strategy and commercial priorities.
• Multi-functional Leadership: Provide hands‑on technical leadership to project teams; set design direction, review architectures and code, and remove blockers on sophisticated problems.
• Agentic AI Systems: Lead the design of multi‑agent systems and LLM applications, including orchestration, tool use, memory, guardrails and human‑in‑the‑loop patterns.
• Evaluation Excellence: Establish standards for evaluation, including automated and human‑in‑the‑loop methods, benchmarking, and quality gates pre‑and post‑release.
• Model Optimization: Drive approaches to fine‑tuning, distillation and reinforcement learning (including RL from feedback) to improve quality, cost and latency over off‑the‑shelf models.
• Production Observability: Set standards for telemetry, observability, monitoring, drift detection and cost management for AI systems in production.
• Emerging Tech to Production: Evaluate emerging models, frameworks and methods; run proofs of concept and translate them into robust, production‑grade solutions.
• Enterprise Alignment: Partner with platform, data engineering, architecture, security and responsible‑AI teams to ensure governance, compliance and reuse.
• Partner Communication: Translate complex technical concepts and trade‑offs for senior non‑technical partners and shape compelling use cases for AI investments.
• Talent Uplift and Representation: Mentor senior data scientists and ML engineers, raise the technical bar, and represent our work in internal forums and partner engagements.

Essential Skills/Experience:

Minimum Requirements:

• BS (or equivalent experience) in computer science, mathematics, statistics, engineering, physics, computational sciences or a related quantitative field.
• Extensive experience (typically 12+ years) building and deploying data science and ML solutions, including considerable recent hands‑on experience (typically 2+ years) with LLM‑based and agentic applications in production.
• Demonstrated track record of setting technical direction and leading complex, multi‑functional AI delivery through influence.

Mandatory Technical Skills:

• Generative AI and LLMs: prompt and context engineering, retrieval‑augmented generation, fine‑tuning (including parameter‑efficient methods), distillation, and model selection and cost/latency optimisation.
• Agentic AI: design and delivery of multi‑agent systems using frameworks such as Amazon Bedrock AgentCore (or equivalents such as LangGraph, Strands or AutoGen), including tool integration, orchestration and guardrails.
• Evaluation and observability: LLM and agent evaluation methods and platforms, telemetry, tracing and monitoring of AI systems.
• Machine learning foundations: strong classical ML (supervised and unsupervised methods, statistical modelling) and deep learning, including transformers, using PyTorch or TensorFlow.
• Programming and tools: expert Python; SQL; scikit‑learn; Jupyter; Streamlit or similar for rapid prototyping.
• Cloud and MLOps: AWS (SageMaker, Bedrock), CI/CD for ML, MLflow or equivalent experiment and model tracking, containerisation (Docker).

Desirable Skills/Experience:

• Master's or PhD in machine learning, computer science or a related field. We prioritise demonstrated skills over credentials.
• Hands‑on experience with reinforcement learning, including RLHF, RLAIF or other preference‑optimisation methods.
• Experience with knowledge graphs, NLP, and structured and unstructured data (text and image).
• Experience with data and pipeline tooling such as Airflow, Spark and data‑quality frameworks.
• Pharmaceutical, healthcare or commercial analytics domain experience (for example next‑best‑action, field‑force, marketing or market‑access analytics).
• Publications, patents or open‑source contributions in AI/ML.
• Exceptional technical judgement, balancing innovation with pragmatism, risk and co