Role Overview

Join OpenAI's mission to ensure safe AGI benefits all of humanity as a Software Engineer, Reliability in San Francisco. This senior-level role sits at the heart of our Applied Engineering team, bridging research breakthroughs with production reality. You'll architect the resilient infrastructure powering ChatGPT, DALL-E, and future frontier models serving millions worldwide.

In this high-impact position, you'll champion reliability amidst unprecedented scaling challenges. OpenAI's systems must balance researcher velocity with enterprise-grade stability, handling explosive traffic growth while maintaining sub-second latencies for real-time AI interactions. Your work directly enables safe deployment of transformative technology to consumers and businesses globally.

The Reliability Engineering team operates as force multipliers, empowering 100+ engineers with self-service tooling that accelerates safe iteration. You'll tackle complex distributed systems problems unique to AI infrastructure - GPU orchestration at exabyte scale, real-time inference latency optimization, and chaos resilience for mission-critical services. This role demands deep systems expertise combined with collaborative humility to succeed in our fast-paced, mission-driven culture.

Success metrics include reducing MTTR by 50%, achieving 99.99% uptime across core services, and enabling 10x capacity growth without proportional headcount increases. You'll partner with world-class researchers, product leaders, and fellow engineers to deliver reliable AI at global scale.

Key Responsibilities

As a Software Engineer, Reliability at OpenAI, your impact spans the full reliability spectrum:

• Scalability Architecture: Design infrastructure handling 100x traffic growth, from millions to billions of daily requests across global edge locations.
• Testing Platform Ownership: Build load, chaos, and synthetic testing frameworks used by every development team, ensuring proactive reliability before production.
• Automation Leadership: Create self-service tools eliminating toil, from auto-scaling GPU clusters to one-click rollback capabilities.
• Resource Lifecycle Management: Develop platforms optimizing CPU/GPU/storage utilization across thousands of nodes, driving multi-million dollar efficiency gains.
• Fault Tolerance Engineering: Implement resilient patterns surviving correlated failures, network partitions, and hardware faults at planetary scale.
• SLO/SLI Framework: Establish measurable reliability objectives guiding engineering decisions across research and product teams.
• Cross-Functional Partnership: Collaborate with researchers deploying bleeding-edge models and PMs launching consumer features, ensuring reliability from day zero.
• Incident Leadership: Lead on-call response for critical outages, driving post-mortems that prevent recurrence through systemic improvements.
• Performance Engineering: Systematically identify and eliminate bottlenecks across the stack, from kernel scheduling to API response times.
• Platform Enablement: Build internal developer platforms accelerating safe velocity for 100+ engineering teams.
• Capacity Planning: Forecast infrastructure needs supporting exponential AI capability growth over 12-24 month horizons.
• Safety Integration: Embed reliability practices ensuring safe AI deployment aligns with OpenAI's core safety commitments.
• Tooling Innovation: Pioneer next-generation observability solving AI-specific monitoring challenges like model drift detection.
• Mentorship: Guide junior engineers while learning from OpenAI's world-class reliability leadership.

Qualifications

Technical Excellence (Required):

• 5+ years production SRE/ reliability engineering in hyper-growth environments
• Deep cloud expertise (AWS/GCP) managing 1000+ node clusters
• Advanced container orchestration (Kubernetes) at massive scale
• Strong systems programming (Python/Go/C++) building mission-critical services
• Proven IaC mastery (Terraform/CloudFormation) across multi-cloud

AI/ML Infrastructure Bonus:

• GPU cluster management experience (NVIDIA/CUDA)
• Experience with Ray, Kubernetes operators for ML workloads
• Familiarity with vector databases, inference serving frameworks

Cultural Fit (Critical):

• End-to-end ownership mindset - runs services soup-to-nuts
• Humble collaborator thriving in flat, high-trust teams
• Relentless curiosity mastering new domains rapidly
• Customer-obsessed serving internal teams as "external" customers

Salary & Benefits

Competitive Compensation: Total cash compensation $220K-$380K+ (base + bonus), substantial equity, comprehensive benefits package.

Exceptional Benefits:

• Top-tier medical/dental/vision with $0 premiums
• 401k with 4%+ match, immediate vesting
• Unlimited PTO - take what you need
• $3,000+ annual learning stipend
• 16+ weeks parental leave
• Fertility benefits, mental health support
• Daily catered meals, gym reimbursement
• Latest hardware, relocation support

Why Join OpenAI?

OpenAI isn't just another tech company - we're building safe artificial general intelligence to benefit humanity. Your reliability engineering directly enables this historic mission.

Unmatched Impact: Every system you build reaches hundreds of millions weekly. ChatGPT alone serves 100M+ users/month.

Technical Challenge: Solve problems no one else faces - exabyte-scale AI infrastructure, sub-100ms global inference, resilient chaos at planetary scale.

World-Class Team: Collaborate with PhDs from Stanford/MIT, ex-Google/DeepMind researchers, builders from Stripe/Uber at their peak.

Mission-Driven Culture: Safety > Growth. We move deliberately, prioritizing long-term correctness over short-term metrics.

Explosive Growth: 10x'd team size in 18 months. Unlimited internal mobility across research, infra, product.

How to Apply

1. Review Fit: Ensure 5+ years reliability experience at scale
2. Prepare Materials: Resume highlighting production impact metrics
3. Submit Application: Click "Apply Now" - include GitHub/portfolio links
4. Interview Process: Recruiter screen → Technical deep-dive → System design → Team match
5. Timeline: 2-4 weeks end-to-end

OpenAI is an equal opportunity employer. We celebrate diversity and are committed to creating an inclusive environment for all employees.