Synthetic Population Simulation Scientist

Aaru; New York (NY); Sur place; USD 120 000 - 190 000; Plein temps

Avantages offerts par ce poste

Visa sponsorship

Relocation support

Equity participation

Résumé du poste

Aaru is hiring a Researcher to design, build, and improve its simulation systems that generate high-fidelity synthetic agents from real-world profiles. You will enable clients to test decisions against realistic populations before committing to them.

The role focuses on grounding agent behavior in real data, using LLMs in hybrid architectures, and evaluating simulation quality against known benchmarks. Strong candidates hold a PhD in a quantitative field and thrive in fast, rigorous research

Qualifications

• Deep experience with synthetic data generation and agent-based modeling.
• Interest in using LLMs as a simulation substrate and producing structured outputs.
• Ability to discuss accuracy vs. distributional fidelity and communicate results.

Responsabilités

• Design and advance synthesis methods for realistic agents from structured and unstructured data.
• Build pipelines delivering population-level distributional accuracy across demographics and behaviors.
• Ground agent behavior in real-world data sources like surveys and datasets.
• Develop hybrid architectures combining LLMs with classical methods from social science.
• Create evaluation frameworks comparing simulations to ground-truth distributions.
• Identify gaps between simulated and observed population behavior and iterate.

Connaissances

Synthetic data generation

Agent-based modeling

Computational social science

Population simulation

LLM-based simulation

Clear scientific communication

Formation

PhD in ML / computational social science / statistics / psychometrics / quantitative field

Outils

None

Description du poste

Aaru is hiring a Researcher to design, build, and improve its simulation systems that generate high-fidelity synthetic agents from real-world profiles. You will enable clients to test decisions against realistic populations before committing to them.

The role focuses on grounding agent behavior in real data, using LLMs in hybrid architectures, and evaluating simulation quality against known benchmarks. Strong candidates hold a PhD in a quantitative field and thrive in fast, rigorous research