ΠΠ»Ρ ΠΌΡΡΡΠ° Ρ ΡΡΠΎΠΉ Π²Π°ΠΊΠ°Π½ΡΠΈΠ΅ΠΉ Π½ΡΠΆΠ΅Π½ Plus
ΠΠΎΠ΄ΠΊΠ»ΡΡΠΈΡΡ Plus
ΠΠΏΠΈΡΠ°Π½ΠΈΠ΅ Π²Π°ΠΊΠ°Π½ΡΠΈΠΈ
Π’Π΅ΠΊΡΡ:
Hirify AI/ΠΡΠΈΠ³ΠΈΠ½Π°Π»
TL;DR
AI Research Scientist (AI/LLM/Psychometrics): Building and validating AI scoring systems for hiring assessments with an accent on LLM evaluation, fairness analysis, psychometric calibration, and predictive performance modeling. Focus on designing human-agreement benchmarks, fine-tuning self-hosted models, detecting demographic bias, and translating research findings into reliable hiring recommendations.
Location: Hybrid in Paris, France; London is also listed as a location.
Company
hirify.global People builds AI-powered hiring tools covering job analysis, assessments, AI interviews, automated scoring, and performance prediction.
What you will do
Evaluate and benchmark closed- and open-source LLMs for automated scoring across assessment types and languages.
Test model accuracy, reliability, consistency, robustness, prompt sensitivity, input noise, and prompt-injection resistance.
Develop self-hosted scoring models, fine-tune open-weight models on expert-rated data, and assess model changes for accuracy, reliability, and fairness.
Design studies comparing AI scores with expert ratings; build hybrid human-and-AI scoring pipelines and calibrate psychometric models.
Analyse demographic bias, apply mitigation methods, and contribute to fairness dashboards and model-governance documentation.
Model links between assessment scores, job performance, and retention; publish research and communicate findings to technical, scientific, business, and client audiences.
Requirements
PhD or MSc in machine learning, computational linguistics, data science, or psychology.
Publication record including peer-reviewed or conference papers.
Hands-on experience evaluating LLMs through benchmark design, human-rater agreement measurement, and consistency and robustness testing.
Experience testing AI-driven decisions for fairness and bias.
Proficiency in Python or R and experience with statistical analysis.
Ability to communicate results clearly to scientific and business audiences.
Nice to have
Experience with NLP techniques for automated scoring.
Experience fine-tuning language models, such as with LoRA.
Experience with cloud databases and large datasets.
Working knowledge of psychometric reliability, validity, IRT, and assessment design and validation.
Working knowledge of ethical AI, data governance, and compliance, including the EU AI Act.
Experience collaborating with engineering, product, and content teams.
Culture & Benefits
Full-time hybrid work in Paris, with London also listed as a location.
Early-employee opportunity to help shape the future of the Science and AI team.
Collaboration with business psychologists, psychometricians, AI scientists, engineering, data science, implementation, and customer success teams.