Leads the team you would join
Signs off on engineering hires
Owns the opening
From your own Gmail and LinkedIn.
About the role
This Bangalore-based senior role sits in AstraZeneca's AI and ML engineering group and centers on generative AI, agentic AI and classical machine learning for pharmaceutical commercial work. The hire leads technical workstreams, writes production code and coaches more junior practitioners while working across commercial, medical and IT teams. The listing was published through Built In, and the work is in-office at least three days a week on average.
What you’ll do
- Create generative AI and agent-based products for commercial use, including multi-agent systems on Amazon Bedrock AgentCore.
- Apply fine-tuning and distillation to lift output quality while lowering cost and latency.
- Maintain evaluation tooling that tracks LLM and agent quality, safety and reliability.
- Instrument production AI with telemetry and tracing, and react to quality, drift and spending issues.
- Assess where reinforcement learning and feedback loops add measurable value.
- Maintain classical ML models for commercial analytics, such as classification, regression, clustering, forecasting and NLP.
- Combine internal and external data with data engineering, informatics and platform teams.
- Guide technical workstreams, mentor colleagues and keep the team current on new methods.
What they’re looking for
- A quantitative bachelor's degree or equivalent practical experience.
- Usually six or more years in data science and ML, with at least a year recently on LLM applications.
- A record of leading AI or data science projects independently.
- Knowledge of prompt and context engineering, RAG, and fine-tuning or distillation.
- Experience building agents or multi-agent flows with AgentCore, LangGraph, Strands or similar.
- Experience designing and running evaluations for LLM and agent systems.
- Classical ML and deep learning with transformers in PyTorch or TensorFlow.
- Very strong Python and SQL, with scikit-learn, Jupyter, Streamlit or similar tools.
- Working knowledge of AWS, with SageMaker and Bedrock preferred.
- Fluent English and the ability to explain methods to non-technical audiences.
Nice to have
- A master's or PhD in a relevant field, although skills are weighed above credentials.
- Experience with preference-based LLM optimization and reinforcement learning.
- Observability tooling for AI systems.
- MLflow, Docker, CI/CD and Airflow.
- Knowledge graph work and experience with text and image data.
- Pharmaceutical, healthcare or commercial analytics background.
What’s on offer
- Hybrid working, averaging at least three office days a week.
- A collaborative, global, cross-functional team setting.
Skills
Questions about this role
When does the posting close?
The closing date is 19 November 2026.
Is office attendance required?
Yes, on average at least three days per week.
How much experience is needed?
Typically six or more years in data science and ML, including at least one year on LLM applications.
Is a PhD required?
No. A master's or PhD is desirable, but demonstrated skills are prioritized over credentials.