Agentic AI Engineer: Build Smart Workflows for Drug Discovery

Healx; Cambridge; Hybrid; GBP 70,000 - 110,000; Full time

Benefits offered by this job

Hybrid working UK-based

Private medical insurance

25 days annual leave

7% employer pension contributions

Life insurance 4x base salary

Personal development budgets

Job summary

Healx in Cambridge is seeking an Agentic AI Engineer to build LLM-agentic workflows that power our drug discovery platform. Reporting to the Director of Tech Strategy, you will design and refine agents that reason over our knowledge graph, literature and methods to generate testable hypotheses for rare diseases.

You'll join a cross-functional team of ML and software engineers, scientists and clinicians, work within an agentic stack (Python, ADK, AgentSpace) and help us adapt as the field evolves

Qualifications

• Experience building LLM-agentic workflows that drive drug discovery.
• Strong Python coding skills with tested, maintainable code.
• Ability to translate complex discovery problems into agent designs.

Responsibilities

• Build and refine AI-agentic workflows on our internal framework to help scientists generate testable hypotheses.
• Translate discovery problems into agent designs and determine where agentic approaches add value.
• Integrate agents with knowledge graph, methods, literature and other sources to reason over evidence.
• Expand and maintain GenAI tools to keep pace with needs.
• Contribute to evaluation of agentic workflows with sensible tests and standards.
• Write clear, maintainable documentation and code.

Tools

AgentSpace

ADK

Knowledge graphs

Job description

Healx in Cambridge is seeking an Agentic AI Engineer to build LLM-agentic workflows that power our drug discovery platform. Reporting to the Director of Tech Strategy, you will design and refine agents that reason over our knowledge graph, literature and methods to generate testable hypotheses for rare diseases.

You'll join a cross-functional team of ML and software engineers, scientists and clinicians, work within an agentic stack (Python, ADK, AgentSpace) and help us adapt as the field evolves