Senior Research Scientist, AI & Workforce Intelligence - Remote, 4-Day Work Week

Wonderlic is seeking a Senior Research Scientist to sit at the precise intersection of I-O psychology and machine learning - someone who has spent time building real systems with real data who cares deeply about what those systems are predicting. This is not a role for an ML engineer who finds people data interesting as a side project, or for an I-O psychologist who has learned to code. We’re looking for someone who has genuinely lived in both worlds and is ready to own the problems that live between them.

Wonderlic built its applied AI/ML team from scratch, developed the Jobs Engine, a first-of-its-kind machine learning system for job analysis at scale, and won the 2025 SIOP Machine Learning Competition. We’re now refining and expanding that work to deliver sharper, richer job analysis insights across thousands of roles. You will own the continued improvement of our jobs engine, the system that ingests labor market data and produces job analysis at scale across thousands of roles, and serve as one of the organization’s experts on how AI should be applied to both employee selection and development.

As Senior Research Scientist, you will set the scientific and technical direction for how Wonderlic understands work at scale and translates assessment science into AI-powered insight. Your work will directly determine the quality, defensibility, and reach of the inferences our platform makes about jobs and people - inferences that drive hiring decisions, development plans, and coaching conversations for millions of users. You will be the person in the organization who can hold both the I-O science question and the production ML question in the same breath, and answer both.

Lead the Continued Development of the Jobs Engine: Own the architecture, integrity, and continuous expansion of Wonderlic’s AI job analysis system, which ingests labor market data from O*NET, LinkedIn, Indeed, and other sources to extrapolate cognitive complexity ratings, norm groups, and occupational interest profiles for thousands of jobs. Build and refine the models that make inferences about work content from unstructured text. Leverage existing occupational taxonomies (e.g., O*NET, ESCO) where appropriate but also expand beyond them. Ensure all outputs are scientifically defensible and scalable as the nature of work evolves.

Act as an Expert on AI Implementation Across the Organization: Partner with product managers, engineers, and I-O psychologists to translate scientific requirements into AI-powered systems and ensure those systems meet the standards the work demands. Advise on which approaches are best suited to specific implementation challenges - assessment interpretation, manager and teams reporting, coaching content - and help teams understand what good looks like before they build and after they ship. Serve as an internal resource on what ML and AI can and cannot do in assessment and organizational contexts, and contribute to Wonderlic’s external scientific credibility.

Drive Scientific Rigor: Apply I-O psychology principles - adverse impact consideration, norm group construction, and evidence-based evaluation standards - to every system you build. Ensure that Wonderlic’s AI products meet professional and legal standards for selection and development tools. Push back when speed is being prioritized at the expense of defensibility, and find pragmatic paths forward when theoretical purity would prevent shipping.

What Success Looks Like: In your first six months, you will have a deep understanding of the jobs engine architecture and shipped at least one meaningful improvement to its coverage, accuracy, or scientific defensibility. You will have established working
relationships with the I-O and Product teams and have a clear roadmap for the continued integration of AI into the platform. 

Within a year, you will have materially expanded the jobs engine’s reach and accuracy with demonstrable improvements to its coverage and the quality of inferences it produces. You will have contributed to the continued integration of AI across the platform, and be the person your colleagues go to with the hardest ML-meets-IO-science problems - not because you have every answer, but because you know how to find them.

The long-term marker of success is a jobs engine that provides defensible, scaled job analysis inference across millions of roles using a wide variety of inputs, AI capabilities integrated into the platform in ways that hold up to scientific and professional scrutiny, and a track record of using these technologies to improve people’s understanding of themselves, their work, and their teams. You will also have helped to develop new assessments that leverage emerging technologies which allow for richer and more secure evaluation of individuals.

Skills & Capabilities:

• Applied NLP and ML engineering skills: embeddings, semantic search, clustering, text classification, transformer architectures, model tuning and evaluation, all on potentially messy, unstructured data
• Occupational data modeling: job titles, task statements, skills, competencies, credentials, job families, seniority levels, title normalization, job similarity, role differentiation, and occupational frameworks such as O*NET and ESCO
• Responsible AI judgment in employment contexts: fairness, explainability, auditability, bias mitigation, human review, and legal and ethical considerations in AI-supported selection and employee development systems
• Generative AI evaluation skills: rubric-based review, groundedness checks, error analysis, regression testing, and evaluation of LLM-generated job descriptions, work-context summaries, and assessment result contextualization.
• Product judgment for applied ML systems: balancing accuracy, explainability, automation, expert review, user input, maintainability, uncertainty, and job-specific nuance.
• Working fluency with assessment and I/O concepts: job relatedness, criterion relationships, adverse impact, norm groups, assessment profiles, and score interpretation.
• Ability to own ambiguous, high-complexity problems: framing underspecified problems, challenging weak assumptions, learning domain constraints quickly, and driving durable solutions in a small-company environment.

Mindset:

• You came to I-O psychology because you care about work - what makes it meaningful, who thrives in it, how to measure fit. That hasn’t changed.
• You have a healthy relationship with “good enough”: you know that perfect is the enemy of shipped, and you have the judgment to know where the line is.
• You can hold both worlds simultaneously: what does this score mean for a real person, and how do I build the system that generates it.
• You are genuinely curious about the problems that exist for both employee selection and development, not just tolerant of them.
• You thrive in an environment that requires creativity and scrappiness: you can work comfortably in a situation where the problems are hard, the team is small, the constraints are many, and the ownership is real.

• Graduate degree in I-O Psychology, Organizational Psychology, Organizational Development, or closely related field (quantitative focus strongly preferred); doctoral degree a plus
• Demonstrated ML engineering experience with shipped, production-grade systems - not just research or coursework
• Experience applying modern NLP methods to behavioral, assessment-based, or labor market data 
• Track record of work that had to be both technically sound and legally/professionally defensible
• Minimum 3 years of applied industry experience; 5+ years preferred
• Experience at the intersection of I-O science and algorithmic fairness strongly preferred
• Familiarity with occupational taxonomies, vocational interests, or cognitive ability frameworks a significant plus

Compensation: 

• $130,000 to $150,000 based on experience and expertise.