Overview
At Lupa, we're hiring an Agentic Platform Engineer for a high-growth B2B SaaS company building the next generation of procurement and practice management (PMS) software for the healthcare market. Already live across 11 markets in Europe with an established client base in Asia, this is a company that operates across time zones, across stacks, and — increasingly — across AI agents.
This is a rare opportunity to help rebuild how software actually ships. The company is moving toward a world where product epics go from spec to production largely through orchestrated AI agents — not just AI-assisted humans typing faster. You'll be the engineer who builds and operates that platform, working directly with the CTO to define what "agent-first" delivery really looks like.
This is not an AI/ML research role. It's an infrastructure engineering role for someone who has spent years in the guts of production systems and has already crossed the line into building and running agentic workflows — someone comfortable giving an agent real autonomy over real infrastructure, and rigorous enough to know exactly where that autonomy should stop.
What You'll Do
Design and operate the infrastructure that lets AI agents build, test, and ship real product work — safely, observably, and at production quality.
Work as an architect: make the calls on stack, tooling, and platform direction as the engineering model evolves, not just execute someone else's design.
Partner directly with the CTO to define and harden the "agent-first" delivery model — this role helps build the playbook, not just follow one.
Own the infrastructure layer end to end: cloud, CI/CD, observability, and the guardrails that keep autonomous agents accountable in production.
Stay stack-agnostic — the platform runs Python, React, and Golang today, but you'll be expected to work across whatever the business needs next.
Who You Are
Must-haves
8–10+ years building and operating production infrastructure — Kubernetes, cloud platforms, CI/CD, distributed systems.
Experience carrying the weight of an architect's decisions, whether or not that was your official title.
A real academic or algorithmic foundation — the kind that lets you question a plan before you run it, catch the failure mode an agent won't see coming, and hold multi-step complexity in your head.
Hands-on experience trying to get AI systems into production — you've done the fighting, not just the reading. (Many of the strongest people for this role come from DevOps and have grown into AI-native work from there.)
Nice-to-haves
Direct experience building or operating multi-agent / agent-orchestration systems (not just using Copilot or Cursor).
Experience in regulated or enterprise-scale environments.
Exposure to healthtech, fintech, or other domains with strong compliance/reliability requirements.
What's In It For You
Compensation: Target range of $5,000–8,000 USD/month, based on experience and seniority.
Location: Fully remote within Europe.
Direct impact: You'll report directly to the CTO and help build the "agent-first" playbook from the ground up — not inherit someone else's.
Growth: Join at a genuine inflection point — the company has 6 years of development behind it, secured funding, committed investors, a major anchor client, and 11 live markets. The hard work is done; this role is about scaling it exponentially.
Application
If you've actually shipped agentic systems to production — and can tell the difference between real autonomy and marketing noise — we want to talk to you.
The process is built to find out, not to take your word for it:
A short take-home project (~1 hour).
A hands-on working session with the CTO (~2 hours), where you'll take the take-home further, live, using AI to get it to production. This is the real interview — and, frankly, the fun part.
Apply through Lupa today and a member of our recruiting team will be in touch about next steps.