Role overview

About The Role

Boson AI builds production-grade AI systems that make communication with AI more natural, capable, and useful. We are looking for a Site Reliability Engineer to help build and operate the infrastructure behind that work.

Based in Toronto or remote, you will work across the systems that enable large-scale AI training and serving: high-performance networks, GPU clusters, storage, scheduling, and the operational tooling that keeps them reliable. This is a hands-on role for someone who enjoys taking complex infrastructure from “it works” to dependable, observable, and scalable.

You do not need to be an expert in every layer of the stack. We are looking for deep strength in at least one area—networking, cluster scheduling, storage, GPU systems, or AI infrastructure— and the curiosity and judgment to collaborate across the rest.

Responsibilities

Design, operate, and improve reliable infrastructure for AI training and inference workloads
Own and automate operational workflows across one or more core areas: networking, compute allocation, storage, GPU/server configuration, or AI platforms
Build monitoring, alerting, runbooks, and incident-response practices that make systems easier to operate
Diagnose performance, capacity, and reliability issues across hardware, operating systems, networks, schedulers, and distributed workloads
Partner closely with ML, research, and platform teams to translate workload needs into practical infrastructure improvements
Improve provisioning, configuration management, testing, and deployment automation
Help plan cluster growth, capacity allocation, upgrades, and lifecycle management
Contribute to a thoughtful reliability culture through documentation, post-incident learning, and pragmatic engineering standards

Minimum Qualifications

4+ years of experience in site reliability engineering, infrastructure engineering, systems engineering, or a related production-operations role
Strong hands-on expertise in at least one of the following:
Networking, including firewalls, switching, routing, ASN/BGP configuration, or InfiniBand
Cluster and systems allocation with Kubernetes, SLURM, MAAS, or similar platforms
Distributed storage, particularly Ceph
GPU and server administration, including CUDA drivers, firmware, BIOS, and hardware troubleshooting
AI training or model-serving infrastructure
Experience operating production systems with a focus on availability, performance, security, and automation
Strong Linux administration and scripting skills
A systematic approach to troubleshooting across multiple layers of a complex system
Clear written and verbal communication skills, including the ability to work effectively with a distributed team

Preferred Qualifications

-
Experience supporting GPU-intensive AI or HPC environments

-
Experience with NVIDIA GPUs, CUDA, NCCL, and high-performance interconnects - Experience with InfiniBand, RDMA, RoCE, or 100Gb+ Ethernet

-
Familiarity with Kubernetes, SLURM, MAAS, Terraform, Ansible, or similar infrastructure tooling

-
Experience operating or tuning Ceph clusters

-
Familiarity with observability tooling such as Prometheus, Grafana, and centralized logging systems

-
Experience with hardware provisioning, firmware management, and bare-metal automation

-
Experience running large-scale distributed training or high-throughput inference workloads

-
Familiarity with cloud and hybrid infrastructure across AWS, GCP, or Azure

$125,000 - $250,000 a yearBoson AI is building AI systems for real-world, business-critical use. If you enjoy solving difficult infrastructure problems and want your work to directly enable the next generation of AI products, we’d love to hear from you.We may use artificial intelligence (AI) tools to support parts of the hiring process, such as reviewing applications, analyzing resumes, or assessing responses and identifying potential inconsistencies or verification signals in application materials based on available information. These tools assist our recruitment team but do not replace human judgment. Final hiring decisions are ultimately made by humans. If you would like more information about how your data is processed, please contact us.