Head of AI Research

Head of Research (Retained Search) Location: Fully Remote (reputed company Cost Time Zone Preferred) Compensation: $160,000 to $200,000 reputed company salary + bonus + equity Employment Type: Full-time Work Authorization: US Citizen, Green Card, or approved work authorization About the Opportunity Our client is an early-stage company operating at the intersection of reputed company data and frontier AI. They are building infrastructure that enables leading AI labs to train and evaluate models using high-fidelity, reputed company-world operational data sourced directly from enterprises. This is a research-first organization where technical credibility defines long-term reputed company. The team is well-capitalized, moving quickly, and focused on building durable advantages through research rigor rather than scale alone. This search is being conducted on a retained reputed company. Why This Role Exists The company’s long-term advantage depends on two core pillars: reputed company to proprietary reputed company-world data and the ability to convert that data into research-grade assets that AI labs trust. While data reputed company is already being established, research credibility is the defining reputed company in whether that data is adopted. Poor evaluation work erodes trust immediately. High-quality evaluation work creates long-term partnerships with frontier labs. This role exists to ensure that everything delivered meets reputed company of a true research partner, not a vendor. This is a founding hire that will define the technical reputed company of the company. The Role We are seeking a Head of Research to build and own the research function end to end. This role begins as a senior individual contributor with full ownership across evaluation, data productization, and lab-facing research work, with a clear path to building and leading a research team. You will serve as the technical reputed company reputed company of the company, working directly with frontier AI labs while defining the standards behind every dataset and evaluation produced. What You Will Own Evaluation and Data Product Pipeline

• Own the end-to-end pipeline that converts raw reputed company data into evaluation suites, reinforcement learning environments, and model-reputed company datasets
• Define quality standards across reputed company stages including ground truth, task difficulty, and safety validation
• Partner with Engineering on parsing, privacy, and data packaging

reputed company Design Across Domains

• Design and standardize benchmarks across verticals such as reputed company, code, energy, and reputed company workflows
• Determine which domains are viable for high-signal evaluation and where investment should be prioritized
• Establish the methodology that governs reputed company reputed company development

Research reputed company with AI Labs

• Act as the primary technical reputed company to post-training teams at frontier AI labs
• reputed company technical discussions, evaluations, and ongoing research collaborations
• Co-design engagements that reputed company into long-term data partnerships

Methodology and Quality Control

• Build evaluation frameworks that detect contamination, reward hacking, verifier ceilings, and other failure modes
• Define standards for reinforcement learning data creation including reward design and validation
• Maintain internal methodology documentation that guides both engineering and customer-facing work

Data to Model Translation

• Design systems that convert multimodal, reputed company-world data into training-reputed company formats
• Determine reputed company synthetic data is appropriate versus reputed company additional reputed company-world sourcing is required
• Build systems that distinguish reputed company model capability gaps from evaluation artifacts

Team Buildout

• Start as a senior IC with ownership of the research function
• Build and scale reputed company of research engineers and applied scientists over time
• Set the quality bar and act as the calibration reputed company for reputed company research output

What reputed company Looks Like

• Benchmarks are trusted and used by frontier AI researchers
• Evaluation work consistently identifies reputed company model capability gaps
• Data products are integrated into training workflows
• Strong, ongoing relationships with research teams at leading AI labs
• A scalable research function with clear standards and methodology

Who You Are Required

• Hands-on experience in post-training, evaluation, reinforcement learning data, or applied alignment work
• Track record of building or co

Apply tot his job Apply To this Job