Staff DevOps Platform Engineer

Location: Remote (Canada) Compensation: $135,000 - $170,000 CAD + bonus + meaningful equity About Orion Digital Orion Digital Corp (NASDAQ/TSX: ORIO) is a publicly listed Canadian fintech operating a multi-engine financial platform spanning lending (Mogo), payments (Carta), and wealth (Intelligent Investing). We have moved past the traditional fintech model. Capital allocation, decisioning, and risk management sit at the core of how the business creates value, and AI runs as an enabling layer across the platform rather than as a standalone product narrative. That layer is real and in use, and the work now is making it faster, safer, and harder to break. Why this role exists This role builds that layer. Platform engineering here is not a support function. Multiple business lines, regulated entities, and a public company's reporting obligations all run on infrastructure this team owns. Every deployment path, permission boundary, and audit trail either makes faster decisioning possible or quietly caps how fast the business can move. Which means the platform is where the strategy succeeds or stalls. If deploying takes a day, if an agent cannot safely touch production, if nobody can tell what changed and why, the enabling layer stays theoretical. We are hiring a Staff Platform Engineer to build it properly. The shift we are hiring for Most platform teams build for human developers. Ours builds for humans and agents, across multiple business lines and regulated entities. That changes the job: Golden paths become machine-executable contracts, not wiki pages. An agent has to be able to discover them, invoke them, and verify the resultEvery probabilistic step needs a deterministic gate behind it. Tests, policy checks, and schema validation stop being hygiene and become the control system that lets autonomy scale safelyAgents are identities. They need scoped credentials, bounded blast radius, and audit trails that hold up across lending, payments, and securities regulationObservability has to answer a new question. Not just "is the service healthy" but "did the agent do the right thing, and how do we know." If that is already how you think about the problem, keep reading. What you will do Build the substrate for agentic engineering Design and run the execution environments agents work in: sandboxed, reproducible, permissioned, disposableStand up and own our MCP and tool-gateway layer so agents reach infrastructure through governed interfaces instead of ad hoc credentialsTurn CI/CD into a validation loop: fast deterministic gates an agent can retry against until it passes, with humans on the loop rather than in it Own the platform Multi-cloud Kubernetes, infrastructure as code, and GitOps (ArgoCD) delivery. Hands on keyboard, not diagramsDrive infrastructure modernization across our cloud environments, including the unglamorous parts: dependency mapping, and documentation that survives your absenceServe every product line without building a separate platform for each one. Where they genuinely differ, respect it. Where they have drifted for no reason, converge themTreat security and compliance as design constraints from the first commit rather than a review at the end Take toil off the company Push autonomous remediation into the paths that page people today. The target is alerts that resolve themselves and incidents where the first responder is an agent arriving with a hypothesis already testedOwn the numbers: deploy frequency, lead time, change failure rate, MTTR, and the share of operational work running without a human in the middle. Move them quarter over quarterWork outside engineering too. Operations, compliance, and support are full of repetitive work a well-built platform can absorb What you need Non-negotiable Deep, current production experience with Kubernetes, Terraform, and a modern GitOps deployment stack. You have owned what you built, including running it in production rather than handing it off once it shippedReal cloud depth. We run across AWS and OCI, so experience in both is an advantage, but what matters is genuine production depth in at least one and the appetite to get fluent in the other quicklyYou run coding and ops agents as part of your daily work, and you can show us. Sustained real use on real systems (Claude Code, Cursor, Codex, or equivalent, plus whatever you have wired together yourself), not a trial and an opinion. Be ready to walk through something you shipped this month where an agent did most of the typing and you did the judgingStrong opinions about where autonomy belongs and where it does not, informed by having been burned at least onceWriting clear enough to be read by a junior engineer and a model Advantage You have built internal developer platforms, MCP servers, or agent tooling that other engineers actually adoptedNon-human identity, secrets architecture, or policy-as-code in a regulated environmentRe-platforming at scale, especially where the old system could not go downOpinions about how to evaluate infrastructure agents. Almost nobody has these yet We are not screening on years of experience as a proxy for capability. We will look at what you have built and how fast you build. How we work Small teams, high ownership, short distance between decision and deploymentWe would rather collaborate than dictate, and we know when to do it well versus when to do it quicklyOutput is the measure Benefits of working with us: Fully Remote Work - Unless otherwise specified in the job posting, Orion offers flexible remote work, supported by the tools and resources needed to succeed. Comprehensive Health Benefits - Access to medical, dental, and vision coverage plus a Health Spending AccountStock Options - Meaningful equity participation aligned with long-term value creation and company growthFlex Vacation - Benefit from paid time off, including vacation and personal daysWellbeing Programs - Access counselling services, mental health support, and additional wellness resources Powered by JazzHR gyQpzvXuwD