Site Reliability Engineer

Site Reliability Engineer Work Model: Hybrid Location: Toronto, Ontario Job Type: Full Time ABOUT THE ROLE That changes what reliability engineering means here, and it makes the role more interesting rather than less. The surface RBC owns the Content Delivery Network (CDN) and Edge – origin configuration, WAF, BOT management, TLS, DNS and cache invalidation – together with the release pipeline, the integrations on the page, and the observability practice that has to be built from real-user telemetry and edge logs rather than a traditional server logs. On this platform, page performance is not a quality attribute; it is the product. You will treat it as a service level objective with the same rigor most organizations reserve for uptime.You will be the reliability engineer for a public-facing banking website in a regulated environment, working across Adobe vendor, RBC InfoSec, Network and CDN, Privacy, Production Support and a lean full-stack development team. WHAT YOU WILL OWN Application Support & Incident Management Own end-to-end monitoring of the RBC-controlled surface: CDN and edge configuration, DNS, certificates, cache and invalidation health, and every third-party integration on the page – Search, Consent Management, Analytics, Personalization and AI services and many more to come.Build and run synthetic monitoring from outside the bank network, per template, per language, because internal-only monitoring cannot see the CDN, DNS and certificate failures this architecture is most exposed to.Run smoke testing of dependent interfaces on every change and maintain the automation packs that do it.Participate in the shared on-call rotation as the platform’s subject-matter escalation, and lead incident management for customer-facing eventsOwn the vendor's escalation path into Adobe: severity mapping between Adobe and RBC incident scales, named contacts, evidence capture, and holding the vendor to its commitment during an event.Handle a class of incident that does not exist on traditional platforms – content published but not visible, invalidation failure, and authoring-source outages – and makes those diagnosable by the service desk rather than by you.Change and Release reliability: Design and operate change management for the platform where the Git repository is production: reconcile a merge-to-main deployment model with RBC change control, so that every production change carries an approved record with stalling delivery.Owns the release pipeline as a production control – branch protection, required checks, lint, performance, and secret-screening gates – and the evidence that they are enforced.Own rollback: revert, republish and purge, rehearsed end to end with a measured recovery time and a named authority who can call it without convening a meetingTreat content publishing as a routine process: hundreds of production changes made by content authors, needing approval evidence, attribution and retention trailRepresent the platform at change advisory board, and own the freeze calendar interaction and release notesBusiness Continuity and Resilience Own the recovery obligation that sits with RBC. Adobe operates delivery resiliently, but customers restore their own content from source version history rather than vendor backups – so the content source, the Git repository and the CDN configuration are RBC’s recovery surface, each needs a tested restore.Hold CDN and edge configuration as code so that a lost or corrupted property is a redeploy rather than an outage with no runbookDefine RTO and RPO with the business against the application criticality tier, document the DR exercise plan, and execute the testing – including failure modes you can actually cause: certificate expiry, invalidation failure, WAF misconfiguration, content source unavailability, and repository compromise.Maintain the operational resilience evidence a regulator expects for a material third-party technology arrangement, and keep the platform exit and portability plan current.Reliability & Performance Engineering Set and defend service level objectives for both availability and page performance. Define Core Web Vitals thresholds per template, run them on an error budget, and report against them.Build the observability practice from the telemetry that exists; real user monitoring on the production domains, CDN access logs streamed to enterprise SIEM as the log source of record, and external synthetics. There is no origin server log – designing around that constraint is part of the job.Own third-party scripts and tag governance as a reliability control. Tags are the dominant cause of performance regressions and are added by teams outside engineering change control; you will define the approval route, measure each tag’s cost adn enforce the budget.Own capacity and cost where they still exist: CDN egress, asset storage and processing, media delivery and any RBC-hosted APIs behind the page. Capacity planning here is a financial operations discipline, not a server-sized one.Publish the reliability and performance reporting that the business, risk and technology leadership use.Compliance and Control Evidence Evidence controls on a platform RBC does not operate – which is harder than evidencing your own, and is where a meaningful share of the role’s effort sitsOwn log ingestion into SIEM with the agreed retention, access recertification across the repository, Adobe Admin Console, content source and CDN, and the audit evidence packSupport privacy, operational risks, control assessment and third-party risk processes with operational evidence and maintain alignment to regulatory expectations for technology, cyber and third-party risks.Keep the configuration management database, support model and assignment groups accurate as the platform estate grows. WHAT DO YOU NEED TO SUCCEED? Substantial hands-on experience operating a high-traffic public website behind an enterprise content delivery network. Depth in CDN configuration – origin and cache behaviour, invalidation, edge logic, TLS and DNS – is the single most important qualification. Akamai and Cloudflare experience is an advantage.Practical web application firewall experience, including tuning false positives against production-like traffic before enforcement, and bot management that protects the site without blocking the crawlers you need.A real observability practice: defining service level objectives and error budgets, and building monitoring from log, real-user and synthetic sources rather than from an agent on a serverWeb performance engineering – Core Web Vitals, load and rendering behaviour, and the ability to read a waterfall and attribute a regression to a specific script.Comfortable with front-end technology: This platform ships JavaScript and CSS to the browser with no server tier; you cannot reason about its reliability without reading and understanding itGit-based release engineering and CI/CD as a production control, including infrastructure and configuration as codeIncident command on customer facing services, and the discipline to produce evidence during an event, not after it.Working effectively in a regulated environment – change control, audit evidence, access management and third-party risk – without treating it as an obstacle. Nice-to-have: Experience operating a vendor-run or SaaS-delivered platform, where reliability means instrumenting, escalating and holding a supplier accountable rather than fixing the tier yourselfAdobe Experience Manager exposure, particularly Edge Delivery Services and Assets as a Cloud ServiceFinancial services or another regulated sectorBilingual delivery – operating a site that must meet the same standard in English and FrenchAccessibility and Search Engine Optimization literally sufficient to recognize when a reliability decision creates a compliance or discoverability problemAutomation in Python, Java or JavaScript, and a preference for encoding a runbook rather than writing one.