Principal AI Engineer

Principal AI Engineer

The Department of Biomedical Informatics (DBMI) at Columbia University Irving Medical Center is accelerating the safe and rigorous use of AI in research and healthcare delivery. We build agentic AI solutions and contribute to the core capabilities of agentic AI platforms. We have an immediate opening for a Principal AI Engineer to help lead this work. The ideal candidate is a talented and self-motivated individual with a strong machine learning background, deep experience with agentic systems, solid programming skills, excellent collaboration skills, and a passion for AI innovation and its dissemination.

The incumbent will work under the direction of a faculty member and closely with researchers and programmers in the department and across the school.

Responsibilities

The Principal AI Engineer will architect, build, and maintain scalable system-level solutions of individual agents, tools, their orchestration and integration into an agentic platform, enabling innovation across clinical research and patient care.

• Define and drive the technical vision and standards for AI/ML and agentic systems guiding their development, evaluation, and responsible adoption at scale
• Partner with clinical, research, and technical stakeholders to align agent requirements, implementation decisions, and data-use expectations
• Lead technical design and architecture for clinical research and patient care agent orchestration, tool integration, memory/context management, and evaluation infrastructure
• Design and build research agents and agentic workflows in collaboration with faculty and research staff

Establish and promote practices for developing, evaluating, and operationalizing AI/ML and agentic systems that ensure quality, scalability, reliability, observability, and maintainability

• Write and review code for critical platform components, including LLM integrations, retrieval systems, and agent tooling
• Evaluate and integrate emerging frameworks, models, and APIs (e.g., LangGraph, LLM providers, MCP servers, orchestration frameworks) into the platform
• Mentor junior and senior engineers/software developers and promote best practices in agentic AI development
• Present and educate the institutional developer and research community on agentic AI and software development best practices
• Translate research needs into technical agent designs and platform features, taking into account risks, dependencies, and tradeoffs
• Additional related duties as requested

Minimum Qualifications

• Master's degree and 5 years' experience with education and training in informatics, computer science, data science, or a relevant field, including experience developing AI/ML solutions
• Demonstrated experience building applications with LLMs (e.g., prompting, fine-tuning, RAG, or API integration)

Preferred Qualifications

• PhD degree and 7 years of experience
• Hands-on experience designing or building agentic AI systems (multi-agent orchestration, tool use, autonomous workflows)
• Familiarity with agent frameworks and protocols
• Experience with clinical data

Other Requirements

• Must have excellent organizational skills with attention to detail, ability to multi-task, and ability to work under deadlines in a fast-paced environment
• The position requires exceptional programming skills, great oral and written communication skills, interpersonal skills, and experience and comfort working independently

Equal Opportunity Employer / Disability / Veteran

Columbia University is committed to the hiring of qualified local residents.