Head of Research

Head of Research (Retained Search)

Location: Fully Remote (West Cost Time Zone Preferred)

Compensation: $160,000 to $200,000 base salary + bonus + equity

Employment Type: Full-time

Work Authorization: US Citizen, Green Card, or approved work authorization

Our client is an early-stage company operating at the intersection of enterprise data and frontier AI. They are building infrastructure that enables leading AI labs to train and evaluate models using high-fidelity, real-world operational data sourced directly from enterprises.

This is a research-first organization where technical credibility defines long-term success. 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 basis.

The company’s long-term advantage depends on two core pillars: access to proprietary real-world data and the ability to convert that data into research-grade assets that AI labs trust.

While data access is already being established, research credibility is the defining factor 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 the standard of a true research partner, not a vendor.

This is a founding hire that will define the technical reputation of the company.

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 front door of the company, working directly with frontier AI labs while defining the standards behind every dataset and evaluation produced.

Evaluation and Data Product Pipeline

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

Benchmark Design Across Domains

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

Research Interface with AI Labs

• Act as the primary technical counterpart to post-training teams at frontier AI labs
• Lead technical discussions, evaluations, and ongoing research collaborations
• Co-design engagements that evolve 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, real-world data into training-ready formats
• Determine when synthetic data is appropriate versus when additional real-world sourcing is required
• Build systems that distinguish real model capability gaps from evaluation artifacts

Team Buildout

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

• Benchmarks are trusted and used by frontier AI researchers
• Evaluation work consistently identifies real 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

Required

• Hands-on experience in post-training, evaluation, reinforcement learning data, or applied alignment work
• Track record of building or contributing to benchmarks used in real research environments
• Deep understanding of evaluation failure modes such as contamination, reward hacking, and distribution shift
• Experience working with modern model harnesses and agentic systems
• Comfortable working with messy, real-world data and converting it into structured outputs
• Strong written communication and ability to produce rigorous technical documentation
• Ability to operate in a fast-moving, high-ownership environment

Preferred

• Published or publicly recognized work in evaluation, reinforcement learning, or post-training

Interested in opportunities specifically in United State? Discover our dedicated Visa Sponsorship Jobs in United State page featuring roles from top employers in this location.

• PhD in Machine Learning, Computer Science, Statistics, or related field, or equivalent experience
• Experience building reinforcement learning environments or evaluation systems for complex domains
• Background in regulated or high-stakes industries such as healthcare, finance, or energy

Bonus

• Prior experience as a founding research hire
• Existing relationships with researchers at frontier AI labs
• Contributions to open-source evaluation or reinforcement learning tools

• Base salary: $160,000 to $200,000
• Performance-based bonus structure
• Meaningful equity aligned with a founding-level hire
• Clear path to building and leading the research function

• Fully remote with US time zone alignment preferred
• Regular in-person collaboration sessions and team off-sites
• Small team with direct access to leadership
• Fast-moving, execution-focused environment

• Initial conversation with leadership
• Deep technical discussion with a senior research advisor
• Take-home methodology design exercise
• Final references