Platform Engineer – AI Evaluation Infra (Go)

arena; San Francisco (CA); Hybrid; USD 170,000 - 230,000; Full time

Benefits offered by this job

Competitive compensation and equity

Health and wellness benefits

Opportunity to work on cutting-edge AI

Small, mission-driven team

Job summary

Arena Intelligence is seeking a Software Engineer - Platform to build core infrastructure for real-world AI model evaluation at scale. You’ll design low-latency APIs, ensure reliability, and enable streaming across providers while collaborating with researchers and product leadership.

This role focuses on building enterprise-grade infrastructure, observability, and developer-friendly APIs, with hands-on work across the stack from gateways to data planes.

Qualifications

• 5+ years of backend engineering experience.
• Strong Go expertise and production backend work.
• Experience with distributed systems and scalable infra.
• Familiarity with LLM API integration and streaming challenges.
• Product-minded with focus on developer experience.

Responsibilities

• Build API-based products from the ground up with low-latency APIs.
• Solve hard streaming problems across heterogeneous providers.
• Ship enterprise-grade infrastructure with security, observability, and cost attribution.
• Build deep observability with tracing, latency breakdowns, and dashboards.
• Collaborate with research to turn ideas into full-featured products.
• Contribute to backend of Leaderboards and Evals platforms when needed.

Skills

Go

Distributed systems

LLM provider APIs

Product mindset

Ambiguity tolerance

Backend engineering

Tools

Kubernetes

Terraform

PostgreSQL

Redis

Job description

Arena Intelligence is seeking a Software Engineer - Platform to build core infrastructure for real-world AI model evaluation at scale. You’ll design low-latency APIs, ensure reliability, and enable streaming across providers while collaborating with researchers and product leadership.

This role focuses on building enterprise-grade infrastructure, observability, and developer-friendly APIs, with hands-on work across the stack from gateways to data planes.