10/05/2026 updated
100 % available
Cloud & Platform Architect · Security · Engineering Leadership · Kubernetes, AWS and GCP, IaC
Vila do Conde, Portugal
Only remote
Training: cloud architecture, containers and Kubernetes, ethical hacking, machine and deep learningAbout me
I design and run cloud platforms: Kubernetes in production, infrastructure as code, and the security and cost work that goes with it. For the last eight years that has been the platform behind a 3D virtual events and learning business.
GitHub ActionsHelmJavaScriptPHPApplication Programming Interfaces (APIs)Amazon Web ServicesAmazon CloudfrontLinuxDevOpsElasticsearchPython (Programming Language)LaravelPostgreSQLMariaDBMongoDB
Cloud Platform Architecture
Kubernetes at production scale (Helm, Skaffold), Nomad on bare metal, AWS (EC2, S3, CloudFront, Lambda, DynamoDB, MediaConvert, SES), Azure (VMs, Entra ID), Terraform/OpenTofu and Ansible, Rook/Ceph, MySQL, MariaDB, PostgreSQL, Redis, Elasticsearch, MongoDB, Kafka, Prometheus, Grafana and Tempo, with cost control tied to the bill.
Real-Time Services & WebRTC
Clustered WebSocket, chat and presence services, WebRTC and conferencing integrations, event streaming, load testing and capacity planning for live events at tens of thousands of concurrent connections.
Engineering Practice & CI/CD
TypeScript, Go, Python, PHP and C; test automation, code review, CI/CD with Jenkins, GitHub Actions, Google Cloud Build and GitLab CI; release management, API-first design, AI tooling and MCP.
Security & Enterprise SSO
Implementation of enterprise single sign-on over SAML2 and OIDC (Microsoft Entra ID, Okta, PingOne), hardening of build and runtime environments, intrusion prevention, and management of findings from external security assessments.
AWS & Azure Media Delivery
Building and running pipelines for video content delivery using S3, MediaConvert to HLS, CloudFront with Lambda functions, and platform email via SES with SQS and SNS; management of EC2, DynamoDB, OpenSearch and Azure VM workloads.
Infrastructure as Code & Bare Metal
Provisioning and deployment of production infrastructure using Terraform/OpenTofu and Ansible on Hetzner bare metal and cloud environments, including Nomad cluster management and multi-region Kubernetes platforms.
Leadership & Team Mentoring
Technical lead of engineering teams, setting engineering practices around testing, code review and CI/CD, mentoring and hiring engineers, co-running onboarding, planning, maintenance and release cadence, and leading enterprise security approvals.
Database Management & Replication
Administration of MySQL and MariaDB with replication, Galera and Percona Cluster, PostgreSQL, Redis, Elasticsearch, MongoDB and Kafka, including migration to operator-managed clusters without client-visible downtime.
AI Tooling & MCP Integration
Building of MCP integration layers, open-source load-testing extensions for real-time services, real-time voice conversation servers, and open-source tooling for AI coding agents and native macOS apps in Swift.
Web Development & Content Platforms
Design and architecture of content platforms with background workers and caching, offline conversion-tracking services across ad-network APIs, A/B testing, rule-based content automation, and fuzzy product search with OCR.
Experience
Professional experience10+ years
Freelancer since1996
Languages
EnglishFluentFrenchBasic knowledgeItalianBasic knowledgePortugueseNative speakerSpanishBasic knowledge
Project history
MootUp is the 3D virtual events product on the shared platform. This role covers the platform's move to the cloud and its real-time services.
- Cloud migration (2020): when the business pivoted to online events and needed high availability and elastic scale, designed and delivered the move from single hosts to a multi-region Kubernetes platform on GKE (Helm, Skaffold, Terraform), with replicated data and clustered real-time services. Live in September 2020; capacity later proven in load tests well beyond the original requirement.
- Real-time services: main author of the platform's signalling service since 2020 (Socket.IO with Redis-backed clustering and autoscaling). It scales horizontally with the cluster and is load-tested to tens of thousands of concurrent users.
- Client events: packaged the chat service (XMPP) for Kubernetes and stood up dedicated clusters for events for Shell, Santander and Axians, with load testing and capacity planning before each.
- Conferencing and video: WebRTC and conferencing integrations (Twilio, Zoom webinars); built and run the pipeline behind clients' video content on AWS (S3, MediaConvert to HLS, CloudFront with Lambda functions).
- Storage and streaming: introduced self-managed Rook/Ceph storage (2021) and ran a Kafka stack for event-driven presence.
LearnBrite is the VR/AR e-learning platform (3D scenarios, a 360° video authoring platform) and the first product on what became the shared platform, alongside Twinity (3D online virtual worlds).
- Secured and optimised the existing server infrastructure and took over its operations and security.
- Built the company's CI/CD on Jenkins with SonarQube (2019) and the DevOps practice every later product runs on: test automation, code review and release management.
- Containerised the applications with Docker and integrated conferencing and streaming services (Twilio, Zoom, Wowza, BlueJeans, Cisco WebEx) into the 3D training platform.
- Modernised a legacy product onto a current framework and schema.
- Took over an acquired product's Azure VMs and moved its Django and PostgreSQL stack into Docker containers (2019).
- Platform email on Amazon SES with SQS and SNS (2020).
- Twinity: supported the server infrastructure and its deployment tooling.
Technical lead for the engineering behind Hyperspace, MootUp and LearnBrite, one shared platform serving clients across pharma, finance, energy, technology and government (Shell, L'Oréal, S&P Global, Novartis, Santander, Siemens, the US State Department), and for ResponsiveVoice. Architecture and feasibility decisions come to this role, from question to production.
- Leadership: set the engineering practice (testing, code review, CI/CD) and mentor the team on it; co-ran developer and QA hiring, including the interview process and code tests; run the planning, maintenance and release cadence and the enterprise security approvals.
- Platform: three GKE clusters defined in Terraform, with self-managed Rook/Ceph storage, database operators for MariaDB, PostgreSQL and MongoDB, and Kafka, Elasticsearch and Redis. Took Ceph across a decade of versions and moved the production databases to operator-managed clusters, all without client-visible downtime.
- Security and availability: over 99.9% measured uptime. Enterprise single sign-on over SAML2 and OIDC (Microsoft Entra ID, Okta, PingOne) for clients including Shell, S&P Global, Invesco and Thermo Fisher: authored the integrations and the setup guides their IT and security teams follow, and worked each rollout through with them. Hardened the build and runtime, added intrusion prevention, and closed every finding from external security assessments.