Frontier AI Researcher & Evaluation Engineer

redcellpartners; USA; Vor Ort; USD 150.000 - 230.000; Vollzeit

Zusammenfassung

Red Cell Partners is launching a stealth AI initiative seeking a Member of Technical Staff, AI Researcher to translate complex workflows into evaluation frameworks and reinforcement learning environments. You will design reward models, verifiers, and controlled experiments to measure and improve model behavior.

This blended research/engineering role operates at the intersection of scientific inquiry and software development, building scalable infrastructure to test ideas and advance frontier AI

Qualifikationen

• 5+ years in ML/AI, CS, data science, or related field.
• Experience bridging AI research and software engineering.
• Strong Python and scalable software development experience.
• Hands-on with AI agents, RL, evaluations, and post-training systems.
• Familiarity with frontier AI methods, benchmarks, and dataset design.
• Experience building reward models and model-feedback systems.
• Design evaluations measuring capability, generalization, alignment.
• Turn research prototypes into production-grade systems.
• Strong analytical and technical writing skills.
• US citizenship or work authorization.

Aufgaben

• Design and build evaluation frameworks, benchmarks, and methodologies for frontier LLMs, AI agents, multimodal models, and RL environments.
• Develop reward models, programmatic verifiers, graders, and other feedback systems that measure and improve model behavior.
• Design and run controlled supervised fine-tuning and RL experiments to study datasets' effects on capability and alignment.
• Research and define what makes evaluations representative of real-world work and environments.
• Build systems for high-quality human data, including task design, annotation methodology, data-quality signals, and data attribution.
• Translate research questions into lightweight, rigorous experiments that yield actionable findings.
• Prototype quickly, evaluate results, diagnose failures, and convert findings into improved benchmarks or systems.
• Move promising methods from experimentation to scalable, production-quality implementations.
• Collaborate across research, engineering, product, and operations to advance the roadmap.
• Contribute to open benchmarks, tools, and technical writing.

Kenntnisse

5+ years ML/AI

AI research & software eng

Python programming

AI agents & RL

Frontier AI methods

Model feedback systems

Evaluation design

Prototype to production

Analytical & writing

US citizen/work authorization

Jobbeschreibung

Red Cell Partners is launching a stealth AI initiative seeking a Member of Technical Staff, AI Researcher to translate complex workflows into evaluation frameworks and reinforcement learning environments. You will design reward models, verifiers, and controlled experiments to measure and improve model behavior.

This blended research/engineering role operates at the intersection of scientific inquiry and software development, building scalable infrastructure to test ideas and advance frontier AI