Lead Applied AI Scientist

ARI is looking for a Lead Applied AI Scientist to build and ship AI methods that make sense of federal research-funding and related administrative data including strategic plans, grants, publications, patents, clinical trials, investigators, institutions, and outcomes. Our clients include the National Institutes of Health. 

This is primarily a hands-on builder role working within a small multidisciplinary team made up of scientists, analysts, economists, and data scientists – all with years of experience working with NIH data. You'll spend the majority of your time designing, prototyping, benchmarking, and productionizing AI methods such as LLMs, NLP, embeddings, retrieval, RAG/GraphRAG, knowledge graphs, and agentic workflow, all applied to real, messy federal research-administration data. Alongside that core work, you'll help set ARI's standards for making these methods trustworthy and defensible in a federal context, and you'll provide technical guidance and supervision to a data scientist. 

We're looking for someone who loves building things that work, who is skeptical of hype, and who can prove, to a federal client's standard, that a method actually holds up. 

What you'll do 

Applied AI Development (the core of this role) 

• Design and build AI methods for analyzing administrative data related to scientific research portfolios.  

• Translate scientific, analytical, and client requirements into testable technical approaches.  

• Build prototypes and production systems using LLMs, NLP, embeddings, semantic retrieval, RAG/GraphRAG, knowledge graphs, and agentic workflows for extraction, classification, entity resolution, retrieval, and synthesis.  

• Benchmark AI methods against simpler analytical approaches, established ML methods, and trained human reviewers to determine real gains in accuracy, reliability, efficiency, or cost.  

• Move validated methods from exploratory testing into robust, maintainable production, working closely with our engineering partners. 

Trustworthy, Reproducible & Explainable AI (secondary, but essential) 

• Help define evaluation standards including reproducibility, explainability, robustness, bias, hallucination risk, scaled to the risk of each application.  

• Design workflows so results trace back to source data, methods, and prompts, and can be reproduced by another analyst.  

• Ensure outputs are explainable to end users: what evidence was used, what the method's limitations are, and where human judgment is required.  

• Contribute to validation reports and documentation for internal review and client assessment. 

Technical Guidance & Continuous Improvement (secondary, but important) 

• Provide day-to-day technical input on ARI's AI methods workstream and mentor other data scientists through code review and methodological guidance.  

• Track developments in applied AI and retrieval/knowledge-graph methods; run focused experiments to decide what ARI should adopt, monitor, or reject.  

• Contribute to proposals, demos, white papers, and publications as needed. 

Qualifications 

Required 

• Advanced degree in computer science, machine learning, data science, biomedical informatics, statistics, or related field - or equivalent experience.  

• Substantial experience applying ML/AI methods to complex, real-world problems, with strong Python skills and a track record of building well-tested, documented, reproducible analytical software.  

• Solid grounding in ML fundamentals: experimental design, model evaluation, uncertainty, error analysis, including setting evaluation criteria and benchmarking against meaningful baselines.  

• Practical experience with LLMs, NLP, embeddings, semantic retrieval, or RAG.  

• Experience with reproducibility, explainability, provenance, auditability, or other responsible-AI requirements.  

• Ability to identify model limitations and communicate them candidly to both technical and non-technical audiences.  

• Ability to independently verify code, methods, and outputs produced with generative AI tools.  

• Ability to lead a technically complex project end-to-end, from problem definition through validation and delivery.  

• Strong collaboration skills and genuine interest in mentoring other staff. 

Nice to Have 

• Experience with knowledge graphs, graph databases, GraphRAG, entity resolution, or linked-data systems.  

• Experience designing or evaluating agentic AI workflows.  

• Experience moving research methods into secure, maintainable production systems, including code review and technical mentorship. 

• Experience with biomedical, scientific, research-administration, or government data, ideally in a regulated or evidence-intensive environment.  

• Familiarity with federal guidance or recognized frameworks for trustworthy/responsible AI.  

• Experience producing model cards, validation reports, risk assessments, or similar governance documentation.  

Position details 

• Full-time, remote within the United States; preference for the DC-MD-VA area. 

• Salary range: $140,000–$170,000, depending on qualifications and experience. 

• Candidates must be authorized to work in the United States and will be required to complete a background check.  

Benefits 

• Employer-paid health and dental insurance for employees with the option to enroll eligible dependents at the employee's expense 

• Employer HSA contributions of $250 per month for employees enrolled in an eligible plan 

• Employer-paid life, short-term disability, and long-term disability insurance 

• Up to 20 days of paid time off a year 

• 11 paid federal holidays a year 

• A 401(k) employer match 

• Profit sharing 

Application materials 

Please submit a resume and a brief cover letter explaining your interest in applying AI to biomedical research and research-policy problems. 

Equal Employment Opportunity Policy: The Analytics Research Institute, LLC, provides equal employment opportunities to all employees and applicants for employment and prohibits discrimination and harassment of any type without regard to race, color, religion, age, sex, national origin, disability status, genetics, protected veteran status, sexual orientation, gender identity or expression, or any other characteristic protected by federal, state or local laws. This policy applies to all terms and conditions of employment, including recruiting, hiring, placement, promotion, termination, layoff, recall, transfer, leaves of absence, compensation and training.