Cassidy Marble

Staff Platform Engineer and Backstage subject-matter expert with 10+ years of experience engineering Internal Developer Platforms (IDPs), high-availability Kubernetes infrastructure, and enterprise GitOps pipelines. Led the migration of a 106k-line legacy Backstage portal into nine independently released, topic-scoped plugin repositories, retiring a 34.8k-line application monolith, raising automated test coverage from a ~32% baseline to an enforced 80% statement threshold on every ported package, and replacing repeated production outages with continuous GitOps delivery on Azure Kubernetes Service.

• Platforms & Orchestration: Spotify Backstage, Kubernetes (AKS/EKS), Flux CD, Helm, Crossplane, Terraform, Docker, Ansible
• Developer Experience & SRE: Golden Paths, IDP Scorecards, Internal Artifact Distribution, Observability (Grafana, Prometheus, Dynatrace)
• Architecture & Security: Azure Workload Identity, External Secrets Operator (ESO), OAuth2/OIDC, Zero-Trust CI/CD, CyberArk
• Programming & Tooling: TypeScript/JavaScript, Go, Rust, Python, Java, Groovy, Bash, GitOps, GitHub Actions, ARC

BlankFactor (now Globant) — Team Lead & Backstage SME

May 2026 – Present · contracted to Finastra

• Contracted to Finastra as Team Lead and Backstage subject-matter expert for the internal developer platform workstream, from initial onboarding through production go-live.
• Led the migration of a 106k-line legacy Backstage IDP portal — 24 custom plugins plus a 34.8k-line application monolith — into nine stand-alone, topic-scoped plugin repositories, each versioned and released independently.
• Replaced Business-Unit semantics with a GitHub-organization entity model (organization = catalog namespace), splitting a 14.5k-line monolithic github-backend into dedicated backend, catalog-module and frontend packages, and dispositioning all 27 legacy endpoints in a written porting map before code moved.
• Remediated GitHub API rate limiting with configurable request timeouts, cache-service-backed responses, a rate-limit monitor and fail-loud credential errors; root-caused and fixed a cross-user cache leak found along the way.
• Raised automated test coverage from a ~32% legacy baseline to an enforced 80% statement threshold on every ported package (~1,900 tests), gated in CI and enforced on every pull request.
• Promoted the plugin fleet from 0.x to stable 1.0.0 in a coordinated seven-repository release wave, then continued shipping minor and major releases as the platform grew.
• Deployed an in-cluster Verdaccio npm registry through Flux GitOps as the sole distribution channel, and root-caused its authentication limitation from the shipped source — token minting is impossible with registration disabled — resolving it with an HTTP-basic identity rolled out to every repository.
• Provisioned self-hosted Actions Runner Controller (ARC) scale sets, one per repository, with a custom runner image built in-cluster with kaniko; defended the runner custom resources against a foreign cluster-wide ARC controller using a Kubernetes ValidatingAdmissionPolicy.
• Implemented two-tier, organization-scoped RBAC — a config-declared admin team with full access, everyone else limited to scaffolder, read-only plugins and catalog reads filtered by organization membership or entity ownership.
• Delivered secrets through Azure workload identity federation and the External Secrets Operator, so no secret value exists in git, in manifests or in CI-rendered output.
• Maintained a 71-decision architecture log across 27 recorded rounds and sustained continuous legacy-parity porting with no code freeze, porting every merged legacy pull request into the plugin repositories.

Nextiva — Senior DevOps & SRE

2021 – 2025

• Operated multi-region Kubernetes clusters and automated Jenkins CI/CD pipelines supporting distributed engineering teams.
• Engineered ML-driven automated support frameworks for real-time threat detection and operational anomaly mitigation.
• Authored private Ansible Galaxy automation collections adhering to GitOps principles to standardize multi-cloud provisioning.

Menard, Inc — Senior Software & Site Reliability Engineer

2017 – 2020

• Executed zero-downtime Redis Cluster migration, driving a 15% improvement in e-commerce page load speeds under peak loads.
• Migrated monolithic search architecture to Kubernetes microservices, reducing deployment overhead by 60%.
• Configured Dynatrace telemetry and automated failover workflows, cutting search incidents by 70%.

Spectrum Progressive School — IT & Infrastructure Lead

2011 – 2014

• Implemented Linux PXE/Netboot automated provisioning infrastructure, extending hardware lifecycles and saving $30,000+ in hardware costs.

• Author, internal Grafana integration plugin suite (frontend, backend and catalog module) built on Backstage's new frontend and backend systems. It deliberately diverges from the upstream @backstage-community/plugin-grafana by routing every Grafana call through a server-side, strictly read-only backend API, so no Grafana credential or per-host configuration reaches the browser.
• Preparing that Grafana plugin for publication as a community plugin — a task assigned and approved in a platform alignment meeting, with submission to the upstream community-plugins project scheduled after the legacy cutover.
• Author, Platform Migration Decision Framework: 71 recorded architectural decisions across 27 rounds, establishing team-wide GitOps, RBAC and release-governance conventions.