LLM Guardrail Strategy & Governance; Manager​/AD

LLM Guardrail Strategy & Governance (Manager/ AD) Compensation Data

The base salary range for the Manager position is $115,000 to $181,000 and for the Associate Director position is $140,000 to $222,000. The position may be eligible for a role specific discretionary bonus, relocation, and/or other compensation elements. We continuously review market data and may adjust salary ranges as needed in the future. Actual compensation will be based on job-related factors such as skills, experience, and qualifications, and other factors permitted by law.

For an overview of our benefits please  .

As an employee of Boehringer Ingelheim, you will actively contribute to the discovery, development and delivery of our products to our patients and customers. Our global presence provides opportunity for all employees to collaborate internationally, offering visibility and opportunity to directly contribute to the companies' success. We realize that our strength and competitive advantage lie with our people. We support our employees in a number of ways to foster a healthy working environment, meaningful work, mobility, networking and work-life balance.

Our competitive compensation and benefit programs reflect Boehringer Ingelheim's high regard for our employees.

The LLM Guardrail Strategy & Governance (Manager/ AD) is responsible for developing, managing, and enhancing AI/LLM guardrail frameworks to ensure safe, compliant, and high-quality outputs from generative AI models across Omnichannel Excellence. This role monitors validation outputs, collaborates closely with Legal, Compliance, IT, Commercial, and Medical functions, and leads the design of new guardrails as the organization expands LLM use across therapeutic areas (TAs) and functions.

The impact of this role is critical in safeguarding the integrity and reliability of AI-driven processes, directly influencing the organization's ability to innovate while maintaining compliance and ethical standards.

This role is based on-site at our Ridgefield, CT location with hybrid flexibility of 2-3 days per week on-site.

Based on the candidate’s demonstrated experience and competencies, this role may be classified at the Manager level. Should this occur, the following duties and responsibilities will be adjusted accordingly to reflect the requirements of that job level.

Guardrail Creation & Maintenance

• Define, implement, and evolve robust guardrails for Large Language Models (LLMs) used across Medical, Commercial, and other Omnichannel offerings.
• Design domain-specific guardrails aligned with scientific accuracy, regulatory compliance, and brand strategy.
• Collaborate with Legal and Compliance to codify business rules and risk thresholds into LLM guardrail logic.

Monitoring & Testing

• Monitor outputs placed in validation (flagged or blocked by the LLM) and investigate root causes to identify patterns and areas for improvement. This proactive monitoring is crucial in maintaining the integrity of AI outputs and preventing the dissemination of inaccurate or non-compliant information.
• Provide recommendations for guardrail tuning (tightening/loosening) based on the review of quarantined content, ensuring that the guardrails remain effective and adaptive to evolving business needs.
• Establish regular testing cycles and QA protocols to assess the effectiveness and precision of guardrail systems, thereby ensuring continuous improvement and reliability of AI outputs.
• Develop metrics and dashboards to track false positives/negatives and performance trends, providing valuable insights for decision-making and strategic planning.

Cross-Functional Collaboration

• Serve as a central liaison between Legal, Compliance, IT, Commercial, and Medical Affairs to align LLM policies with business strategy. This collaboration ensures that AI initiatives are integrated seamlessly across the organization and support overall business objectives.
• Partner with IT and data science teams to test, implement, and validate technical rule enforcement within LLM workflows, ensuring that the guardrails are technically sound and effectively implemented.
• Educate internal stakeholders on…