Manager, AI Engineering and Vendor Operations

NPR•Columbia, WA

1d•Hybrid

About The Position

NPR's AI Labs team is seeking a Manager of AI Engineering and Vendor Operations to serve on the AI Labs leadership team and help shape NPR's AI technical strategy. This player-coach will supervise and mentor AI engineers, support engineering teams across NPR, and lead the evaluation and deployment of AI tools and vendors. The role calls for sound build-versus-buy judgment, clear communication of technical trade-offs to non-technical colleagues, and accountability to NPR standards for privacy, security, and accuracy. The manager will lead technical planning and delivery for AI Labs engineering work, establish engineering practices, review architecture and code, and build automated testing and model evaluation approaches. The position also coordinates production incident response and follow-up, evaluates AI vendors and tools, supports NPR-wide AI tool review with IT Security, and manages selected vendor relationships and usage. The role is intended to help NPR use AI responsibly in service of its mission.

Requirements

5+ years of experience building and shipping production AI or machine learning systems, with deep hands-on knowledge of large language models and related systems.
2+ years of experience managing or mentoring engineers.
Experience setting team engineering practices, such as code review, testing, or deployment standards.
Experience working with technology vendors and managing software tools, including security and data privacy reviews.
Experience building automated testing, including methods for evaluating AI model output quality.
Hands-on experience building in a major cloud environment: Azure, AWS, or GCP.
Experience working across technical and non-technical teams, explaining technical and cost trade-offs to non-technical people, and working with product managers to gather stakeholder requirements.
Strong Python skills and working knowledge of retrieval-augmented generation (RAG) and semantic search.
Proficiency in AI system design and architecture, and AI evaluation systems.
Proficiency in SQL and cloud data warehouses such as BigQuery or Snowflake.
Familiarity with model evaluation methods, prompt engineering, and API integration, including NPR-approved model families such as Claude, Gemini, Gemma, and Llama.
Ability to assess vendor security documentation, data processing terms, and seat-based and usage-based pricing.
Commitment to data privacy and AI ethics, including identifying and reducing bias.
Bachelor's degree in Computer Science, Software Engineering, or a related field, or equivalent practical experience.

Nice To Haves

Experience in a media organization, newsroom, or public service institution.
Experience onboarding staff into a new or growing team.
Experience working with procurement, legal, or vendor risk teams on technology reviews.

Responsibilities

Manage and mentor entry- and mid-level AI engineers, including hiring, onboarding, and supporting professional growth.
Lead technical planning, scoping, and sprint execution for AI Labs engineering work in partnership with AI Labs leadership and editorial and business stakeholders.
Develop AI coding frameworks, practices, and expectations with AI Labs leadership.
Review designs and code, and contribute hands-on as appropriate to keep systems maintainable, secure, and consistent with NPR technical standards.
Build and maintain automated testing and model evaluation practices, including regression tests and pre-release evaluation checks aligned with product quality bars.
Coordinate responses to production bugs and infrastructure incidents and lead follow-up reviews to prevent recurring problems.
Lead technical evaluations of AI vendors, models, and tools, considering architecture, data handling, security, and cost, and recommend whether NPR should build, buy, or partner.
Participate in AI Labs' review of AI tools requested by NPR teams, working with IT Security through vendor risk and security intake processes.
Manage selected AI vendors and tools day to day, including onboarding, seat provisioning, renewals, and offboarding.
Track vendor usage, spend, and renewal dates against budget, and ensure AI systems scale securely and cost-effectively while protecting data privacy.