ARC Prize Foundation Seeks Platform Engineer to Scale AI Benchmark Infrastructure

The ARC Prize Foundation, a Y Combinator-backed organization founded in 2024, is seeking a senior platform engineer to serve as the technical owner and architect of their benchmark infrastructure. This remote, full-time role offers a salary range of $150,000 to $250,000 for candidates with 6+ years of experience in distributed systems and software architecture.

The foundation's mission centers on creating AI benchmarks that measure general intelligence and inspire new ideas in the field. Their ARC-AGI series has become a significant reference point for evaluating AI systems' capabilities beyond narrow task performance.

The Technical Challenge

The selected candidate will take ownership of the V3 backend and infrastructure, with responsibilities spanning multiple critical areas:

Platform Stabilization and Extension

• Maintaining and improving the current benchmark platform's reliability
• Optimizing performance to handle high-volume model evaluations
• Ensuring production-grade stability for a system used by researchers and AI companies globally

Verification and Testing Infrastructure

• Building automated model run pipelines
• Creating reproducible evaluation systems
• Developing data capture and querying capabilities for deeper model analysis
• Implementing scoring systems that can handle diverse AI architectures

Future Benchmark Development

• Supporting ARC-AGI-4 implementation with new environments and human data collection
• Laying technical groundwork for ARC-AGI-5
• Architecting systems that can evolve with advancing AI capabilities

The role requires expertise in Python backend development, distributed systems, SQL databases, and cloud infrastructure. Experience with AI/ML evaluation systems or high-volume technical platforms is particularly valuable.

Why This Matters

The ARC Prize Foundation operates at a critical intersection in AI development. As models become increasingly capable, traditional benchmarks often fail to capture meaningful differences in general intelligence. The foundation's work provides researchers and companies with tools to understand where AI systems excel and where they fall short.

This position offers the opportunity to shape how the AI community measures progress. The technical decisions made by the platform engineer will directly impact which research directions receive attention and funding, influencing the trajectory of AI development itself.

The Team and Culture

With a team of just four people, the ARC Prize Foundation operates with high agency and technical ownership. The current president, Greg Kamradt, emphasizes direct collaboration with founders of YC-funded startups, suggesting a network-rich environment for someone interested in the broader AI ecosystem.

The remote nature of the role, combined with the foundation's focus on US citizens or visa holders, indicates a commitment to building a cohesive team despite geographical distribution.

Technical Stack and Requirements

While specific technologies aren't detailed in the posting, the requirements suggest a modern, cloud-native infrastructure:

• Backend: Python-based systems for model evaluation and scoring
• Infrastructure: Cloud platforms supporting distributed workloads
• Data: SQL databases for structured benchmark results
• Reliability: Production systems requiring 24/7 availability
• AI/ML: Integration with various model architectures and evaluation methodologies

Candidates should be comfortable acting as both individual contributor and technical lead, given the small team size and the scope of responsibility.

Application Process

Interested candidates can apply directly through the Y Combinator platform. The posting emphasizes connecting "directly with founders of the best YC-funded startups," suggesting a streamlined application process for qualified technical talent.

The role represents a unique opportunity to work on infrastructure that shapes how the AI community understands progress in general intelligence. For engineers passionate about AI evaluation, distributed systems, and building platforms that influence research directions, this position offers both technical challenge and meaningful impact.

Apply to the Platform Engineer - Benchmark Lead role