Software Engineer - Infrastructure (Technical Leadership)

Meta; United States; Remote; USD 250,000 - 350,000; Full time

Job summary

Meta is seeking an experienced Software Engineer to build AI agents for production reliability at scale. You will design, prototype, and ship agentic systems that autonomously investigate incidents, identify root causes, and execute safe mitigation with human supervision.

The role blends large-scale distributed systems, AI tuning, and observability to improve agent accuracy and real-world impact. You will work hands-on from architecture to production deployment at Meta scale.

Qualifications

• Bachelor's degree in Computer Science, Computer Engineering, or related field.
• 12+ years of software engineering experience in distributed or infrastructure systems.
• Experience leading multi-year engineering efforts and setting technical direction.
• Experience applying AI/ML to production problems.
• Proven production deployment experience and measurable impact.
• Experience diagnosing production systems using telemetry, configuration, and dependency information.
• Proficiency in languages like C++, Java, Python, Rust.
• Hands-on with production systems from architecture to implementation.
• Experience influencing senior engineers without direct authority.
• Experience building hyperscale, reliable, low-latency systems.
• Experience improving AI agent quality via eval, RL, fine-tuning, distillation, or model-serving optimization.

Responsibilities

• Define the technical vision and architecture for agentic reliability systems across Meta.
• Personally design, code, and ship production agentic systems for complex infrastructure problems.
• Lead the development of an AI agent for production incident investigation and mitigation.
• Develop improvements in agent reasoning, context, tool use, planning, evaluation, learning, and safe execution.
• Explore techniques like fine-tuning, reinforcement learning, distillation, and pruning for gains.
• Identify high-value AI applications across incident prevention, detection, and observability.
• Move from problems to prototypes, validate with real workloads, scale to production.
• Build evaluation and experimentation systems linking agent quality to outcomes.
• Create closed-loop improvements turning outcomes into evaluations and experiments.
• Establish architectures and guardrails for production actions with oversight.
• Collaborate with Infrastructure, AI, Product, Reliability leaders to integrate agentic capabilities.
• Mentor engineers across distributed systems and applied AI.

Skills

Distributed systems

Incident response

Applied AI

Production engineering

C++, Python

Leadership

Mentoring

Debugging

Education

Bachelor's degree in CS/CE or related field

Job description

About

Meta is seeking an experienced Software Engineer to build the next generation of AI systems for production reliability. This role sits at the intersection of large-scale distributed systems, incident response, and applied AI. You will lead the development of an AI agent that can autonomously investigate production incidents, identify likely root causes, create safe mitigation plans, and execute them while humans supervise and can intervene. You will improve the agent's accuracy, autonomy, and real-world impact, while identifying other opportunities to apply agentic systems across incident prevention, detection, mitigation, observability, and infrastructure operations. This may include improving the agent's architecture, context, and tooling, or adapting foundation models through fine-tuning, reinforcement learning, distillation, pruning, and inference optimization.This is an applied engineering role, not a research position. The ideal candidate combines deep infrastructure expertise with sound judgment about when and how to apply AI to production problems, a long-term technical vision, and the ability to move quickly from an ambitious idea to a production system operating safely at Meta scale.This is also a deeply hands-on senior IC role. You will personally prototype, build, evaluate, and ship AI systems, working directly in the code and with production data. A central challenge is creating systems that improve from experience: learning from outcomes, identifying their own failure modes, testing changes, and safely increasing their effectiveness over time.

Responsibilities

• Define the technical vision and architecture for agentic reliability systems across Meta
• Personally design, code, and ship production agentic systems for complex infrastructure problems
• Lead the development of an AI agent for production incident investigation and mitigation, advancing it toward accurate, trusted, and safely supervised autonomous action
• Develop major improvements in agent reasoning, context, tool use, planning, evaluation, learning, and safe execution
• Explore and apply techniques including automated hill climbing, fine-tuning, reinforcement learning, model routing, distillation, pruning, and inference optimization, selecting the simplest approach that produces measurable gains
• Identify new high-value applications of AI across incident prevention, detection, mitigation, observability, and infrastructure operations
• Move rapidly from ambiguous problems to prototypes, validate them against real production workloads, and develop successful approaches into reliable systems at scale
• Build evaluation and experimentation systems that connect agent quality to outcomes such as investigation accuracy, successful mitigation, incident duration, and reduced operational work
• Build closed-loop improvement systems that turn production outcomes into evaluations, experiments, and better agent behavior
• Establish architectures and guardrails for production actions, including authorization, independent validation, auditability, rollback, and human oversight
• Partner with Infrastructure, AI, Product, and Reliability leaders to integrate agentic capabilities into Meta’s production ecosystem
• Influence technical strategy across organizations and mentor other engineers working on distributed systems and applied AI

Minimum Qualifications

• Bachelor's degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience
• 12+ years of software engineering experience, including experience building and operating large-scale distributed or infrastructure systems
• Experience setting technical direction and leading complex, multi-year engineering efforts across organizational boundaries
• Experience applying AI or machine learning systems to production problems
• Demonstrated experience moving from technical concept to production deployment and measurable impact
• Experience diagnosing complex production systems using telemetry, code, configuration, and dependency information
• Experience coding in languages such as C++, Java, Python, Rust, or equivalent
• Recent hands-on experience building and shipping complex production systems, with the ability to move directly between architecture, experimentation, debugging, and implementation
• Experience influencing senior engineers and leaders without direct organizational authority Experience with autonomous or semi-autonomous production actions and their safety, authorization, and rollback mechanisms
• Experience building systems that operate at hyperscale under strict reliability and latency requirements
• Experience improving agent quality through evaluation, context engineering, fine-tuning, reinforcement learning, distillation, or model-serving optimization
• Track record of identifying unconventional opportunities, rapidly prototyping solutions, and changing the technical direction of a large organization
• Deep expertise in observability, incident response, change safety, distributed systems, or production infrastructure
• Experience building self-improving or self-evolving systems, including automated experimentation, feedback loops, hill climbing, reinforcement learning, or recursive self-improvement
• Experience building production AI agents that reason across telemetry, code, configuration, deployments, and operational knowledge