Senior SaaS and DevOps Engineer - Part-time - Worldwide

About the role

Take ownership of the systems that make the company dependable. We are building a procurement platform for purchasing teams across industries and are looking for a senior engineer who can turn a working product into a secure, recoverable and well-operated SaaS platform.

You will work directly with the founder, who builds product features with Claude Code. Your focus is production readiness and platform ownership: review the risks, make pragmatic improvements and leave the system easier to operate. You will have room to challenge technical decisions and explain what needs to happen before a release is safe.

Your first 90 days

Establish reliable environments. Validate and document separate test, demo and production environments on Railway, each with its own Supabase project in Frankfurt. Make branch-based deployments repeatable and document how a release moves between environments.

Prove that recovery works. Apply SQL migrations in the order test, demo, then production. Enable Point-in-Time Recovery and complete a documented restore drill so recovery is demonstrated, not assumed.

Close the security gaps. Review Supabase Auth, JWT, invite-only access, roles, tenant isolation, uploads and Microsoft 365 mail. Secure secrets, protect GitHub branches and identify any release-blocking security issues.
Make operations observable. Introduce health checks, web and worker logging, actionable alerts and practical outage and hotfix runbooks. Keep test and demo email mocked; configure Microsoft Graph for production and verify exactly one message per supplier.

Ongoing ownership

Review pull requests for migration safety, performance, security and tenant risks. Maintain deployment and recovery procedures, help diagnose incidents and protect the integrity of the immutable audit log. Help select and configure an EU AI provider once the data processing agreement is in place. Record decisions and operational changes in docs/ so knowledge stays inside the company.

Success means predictable releases, tested recovery and security risks surfaced early. The founder should understand how the platform operates. Feature development is a secondary part of this role.

THE STACK AND THE WAY WE WORK

Backend

• Node.js 22, ES modules, custom API server
  Frontend
• Next.js with static export, React, TypeScript
  Data and authentication
• PostgreSQL / Supabase, SQL migrations, PGlite tests; Supabase Auth, JWT, invitation-based access
  Infrastructure
• Docker, Railway web and worker services, TransIP DNS
  Email
• Microsoft 365, Microsoft Graph
  Delivery
• GitHub, pull requests, GitHub Actions, Prettier
  Security
• Tenant isolation, immutable audit log, GDPR

What you bring

• You have 5+ years of engineering experience and have taken at least two web applications into production with real users and real data. You can own deployment, database reliability, security and incident response, and are comfortable working across application code and infrastructure.
• You review AI-assisted code critically, recognise migration and tenant-isolation risks, and explain your decisions in plain English. You work independently, document what matters and raise concerns before they become production problems.
• Useful additional experience. Experience in healthcare, government or finance, familiarity with ISO 27001 or NEN 7510, work alongside AI coding assistants, or ownership of a product in a small team would be valuable.

How we work

• Work from anywhere in the world, whether that is the Netherlands, Bali or somewhere else. We agree a workable time for one fixed one-hour meeting each week, with short messages as needed. Time-zone overlap, availability and incident expectations are agreed upfront within the 8-16 hour engagement.
• English is welcome in meetings, reviews and documentation. Code and commits are in English. Some product copy and existing documents are in Dutch; translation tools are fine.
• Start each change on a branch from develop and submit a pull request with passing formatting, type checks, tests and build. Database changes use new migrations only. Sensitive changes receive a dedicated security review. Production keys and customer data never go into code or an AI assistant. Significant changes start with a short written plan.