AI Automation Engineer at Specter in Antwerp, Antwerp, Belgium

The role

This is not a blank page R&D job. You inherit working prototypes and make them reliable, secure and scalable.

You decide what gets hardened, what gets rebuilt and what gets replaced by a better tool.

You work directly with the founder: they keep prototyping new ideas, you decide what is ready for production.

What you'll do

• Make it reliable: error handling, retries and rate limits, so no workflow fails silently.
• Make it visible: monitoring and alerts, so we know a flow broke before a recruiter or client does.
• Structure the data: one clean data model that is the single source of truth for candidates, clients and vacancies.
• Secure it: credentials out of the workflows, proper access control and GDPR compliant data retention.
• Measure AI quality: test sets for AI output, versioned prompts and cost per run tracking.
• Make it last: version control and documentation, so the system never depends on one person.
• Build, rebuild or buy: decide per workflow, and integrate vendor tools where they beat our own.
• Own the production gate: nothing becomes business critical without your sign off.

What we're looking for

• At least 2 years experience in backend, data or automation engineering, including systems other people relied on every day.
• Strong Python, SQL and REST APIs.
• Hands-on experience with workflow tools like n8n, Make or Zapier, and the judgment to know when a flow belongs in code.
• Experience running LLM APIs in production: prompt management, output validation and cost control.
• You think in failure modes: what breaks at ten times the volume, and what happens when the model returns nonsense.
• You can explain technical tradeoffs to non technical people, and you're comfortable saying "this needs rebuilding".

Nice to have

• Experience with CRMs/ATS
• Familiarity with GDPR and the EU AI Act as they apply to recruitment.
• Cloud experience (Azure, AWS or GCP) and infrastructure as code.