Behavioral AI Scientist - Hybrid, Applied Research

Seldon; Greater London; Hybrid; GBP 110,000 - 145,000; Full time

Benefits offered by this job

28 days PTO

Tower Bridge office

2 days WFH per week

Team dinners and offsites

SF travel opportunities

Job summary

Seldon is seeking a highly skilled researcher to join their Research team in London. You’ll build psychologically grounded models of user behavior and deploy them in live interfaces, shaping agent actions and outcomes.

You’ll work with engineers to productionize your models and run human studies to validate results. Ideal candidates hold an MSc/PhD or postdoc in related fields and have experience with decision-making research, RL or agent-based modeling.

Qualifications

• Candidates should have advanced research in decision-making, learning, or behavioral modeling.
• Experience leading human studies is beneficial.
• Comfort interpreting data and collaborating with engineers in fast-moving environments.

Responsibilities

• Model human differences and build computationally useful models.
• Infer users' internal states from behavior and link to persona.
• Shape behavior in live interfaces and study attention and engagement.
• Measure fidelity by comparing agent behavior to human data and benchmarks.
• Understand user interpretations of interfaces and attention patterns.
• Experiment with synthetic populations to test responses before launch.

Skills

Personality & motivation reasoning

Prototype in ML/LLM sandbox

Collaboration with engineers

Research leadership in digital context

Agent-based modeling / RL experience

UX/experimental tools collaboration

Education

MSc, PhD or postdoc in Cognitive Science/Behavioral Economics/Psychology/HCI

Tools

ML sandbox

Job description

Seldon is seeking a highly skilled researcher to join their Research team in London. You’ll build psychologically grounded models of user behavior and deploy them in live interfaces, shaping agent actions and outcomes.

You’ll work with engineers to productionize your models and run human studies to validate results. Ideal candidates hold an MSc/PhD or postdoc in related fields and have experience with decision-making research, RL or agent-based modeling.