Founding Applied AI Researcher - Model Eval & Data Strategy

Morpheus Talent Solutions; United States; On-site; USD 120,000 - 180,000; Full time

Job summary

Morpheus Talent Solutions is seeking an Applied AI Researcher – Model Evaluation & Data Strategy in San Francisco. This in-person role has potential remote options for exceptional candidates and centers on evaluating models and shaping data programs.

You'll design evaluations for generative, multimodal, and agentic systems, analyze outputs to quantify failure modes, and propose data, rubrics, and workflows to fix them, collaborating with operations and engineering.

Qualifications

• Experience in applied AI research, ML, model evaluation, or data-centric AI; hands-on foundation/generative model evaluation; strong Python plus model APIs and structured datasets; familiarity with SFT, preference optimization, RLHF/RLAIF, reward modeling, synthetic data, or LLM-as-a-judge; strong technical writing; ability to run ambiguous research independently.

Responsibilities

• Design evaluations for generative, multimodal, reasoning, tool-use, and agentic systems.
• Analyze outputs and benchmarks to quantify failure modes.
• Recommend SFT data, preference data, expert demonstrations, critiques, and eval sets.
• Build quality systems (calibration, blind review, adjudication).
• Design pilot studies that prove interventions move performance.
• Publish research that positions the company as a research partner to the field.

Skills

Applied AI research

Model evaluation

Data-centric AI

Python

Structured datasets

Education

PhD in CS/ML

Tools

Python

Model APIs

SFT/RLHF

Job description

Morpheus Talent Solutions is seeking an Applied AI Researcher – Model Evaluation & Data Strategy in San Francisco. This in-person role has potential remote options for exceptional candidates and centers on evaluating models and shaping data programs.

You'll design evaluations for generative, multimodal, and agentic systems, analyze outputs to quantify failure modes, and propose data, rubrics, and workflows to fix them, collaborating with operations and engineering.