AI/ML Engineer

BackOps•San Francisco, CA

About The Position

BackOps is building Relay, an AI resolution platform that deploys AI operators to understand operational workflows, work across existing enterprise systems, take action, navigate exceptions, and carry work through to resolution. The company recently raised a $42M Series B led by Insight Partners and is scaling across logistics, manufacturing, retail, and other industries. The AI/ML Engineer will own the infrastructure and systems that make Relay's intelligence more capable, efficient, and reliable. This senior individual-contributor role sits at the intersection of machine learning, distributed systems, and production AI. The engineer will design infrastructure for model execution, inference optimization, evaluation, and continuous improvement, and help determine how models are selected, adapted, deployed, and optimized for enterprise production. The role involves making foundational technical decisions and building capabilities from the ground up in close collaboration with the CTO and engineering teams. The focus is on production AI/ML infrastructure and systems ownership—not simply integrating AI APIs or maintaining an established ML stack. The engineer will establish and improve the foundation on which BackOps scales its AI capabilities, balancing intelligence and quality with performance, scalability, reliability, and cost.

Requirements

6+ years of relevant software engineering experience, with meaningful ownership of complex technical systems at the Senior or Staff level.
Experience building, deploying, or optimizing production AI/ML infrastructure, model-serving systems, inference pipelines, or similarly sophisticated AI systems.
Exceptional engineering fundamentals and strong proficiency in Python.
Experience designing and operating reliable distributed systems.
Understanding of modern foundation models and tradeoffs in model selection, inference architecture, quality, latency, throughput, scalability, and cost.
Hands-on experience with cloud infrastructure, containerization, orchestration, and the operational challenges of production AI workloads.
Understanding of model evaluation and experimentation, with the ability to connect technical changes to measurable improvements in model quality or system performance.
Ability to work in rapidly evolving technical environments and balance immediate needs with long-term scalability.
Ability to make decisions with incomplete information, explain technical tradeoffs, and adapt as technology and evidence evolve.
High standards for engineering quality while moving with urgency.

Nice To Haves

Experience building foundational AI/ML infrastructure at a breakout AI company or exceptional technology organization where performance, reliability, and scale were critical.
Deep experience with PyTorch, Hugging Face, vLLM, Triton, Ray, Kubernetes, or comparable model-serving and distributed-computing frameworks.
Experience designing systems for model optimization, fine-tuning, post-training, evaluation, or continuous improvement of production AI capabilities.
Experience building infrastructure for agentic AI systems, complex model orchestration, or multi-step autonomous workflows.
A record of measurable improvements in model quality, inference efficiency, infrastructure cost, or production reliability.
Experience building shared technical capabilities from zero and establishing architectures and engineering standards for other teams.
Supply chain or logistics experience.

Responsibilities

Own the architecture and engineering of AI/ML systems that power Relay, enabling reliable, efficient operation at scale.
Build and improve model-serving systems, inference pipelines, and execution infrastructure.
Develop model selection, routing, and orchestration capabilities, as well as shared interfaces for foundation models and specialized ML capabilities.
Create infrastructure for experimentation, evaluation, production deployment, model versioning, configuration, and lifecycle management.
Make pragmatic decisions about building custom infrastructure versus using existing providers.
Optimize latency, throughput, scalability, and inference cost; identify bottlenecks, evaluate architectural tradeoffs, and implement improvements.
Develop systematic approaches to improve model quality and effectiveness, including evaluation, benchmarking, comparative testing, context engineering, retrieval, fine-tuning, and model adaptation where appropriate.
Partner with engineers developing Relay's agent capabilities to identify model limitations that cause execution failures and determine how to address them.
Build production reliability capabilities, including observability, failure detection, recovery mechanisms, performance monitoring, and operational controls.
Establish engineering standards for safely introducing models, evaluating changes, managing infrastructure dependencies, and operating complex AI systems in production.
Establish technical metrics and feedback loops for model quality, inference performance, cost, reliability, evaluation, resource utilization, and deployment speed.
Work with Product and Engineering to connect infrastructure improvements to accurate decisions, reliable execution, and autonomous resolution outcomes.
Move between foundational architecture, hands-on implementation, performance debugging, and production operations.