Backend Engineer - AI-Native Platform Builder

Location and Culture Note

Please note that we only accept candidates living in or near Amsterdam, Netherlands. While we fully support remote work, we also value culture and connection. Therefore, our ideal candidate is enthusiastic about coming to our office at least once a week to share the energy and excitement of what we are building together.

Backend Engineer - AI-Native Platform Builder

The Elevator Pitch

Huub is building an AI-driven platform that fully automates complex business processes. No legacy systems with a chatbot glued on. No manual workflows with an AI sauce. We build the infrastructure on which AI agents run, scale, and remain reliable — from the first request to production at scale.

We are looking for a Backend Engineer who understands that building software in 2026 is fundamentally different from two years ago. Not someone who "also uses Copilot sometimes." Someone who leverages Claude Code, MCP servers, skills, and plugins as primary tools — and thus builds in a week what a traditional team takes a month to do.

What you will do

• Build a platform for AI agents. Our agents need a reliable, fast, and scalable backbone. You design and build the services, APIs, and data pipelines that make this possible. Event-driven architecture, async processing, real-time communication — you ensure the foundation is solid.
• Develop AI-natively, every day. You don't write everything yourself, line by line. You orchestrate. You use Claude Code as your primary development environment — with custom skills, MCP integrations, and plugins that accelerate your workflow. You build and share tooling that makes the entire team more productive. You understand that the best engineer is not the fastest typist, but the one who uses AI most intelligently.
• Design and build MCP integrations. Our agents are connected to external data sources and internal tools via the Model Context Protocol. You build and maintain these server integrations — reliably, quickly, and with clear contracts. From third-party data to internal calculation modules: you make them available as tools that agents can invoke.
• Build systems that make AI agents reliable. Observability, tracing, error handling, audit trails — everything needed to run agents in production in a regulated domain. You don't just build features; you build the confidence that those features work correctly.
• Iteratively improve in production. No semi-annual releases. We ship continuously. You work with the product and AI team to improve systems based on real data and user feedback. What you conceive on Monday can be live on Thursday.
• Mentor other engineers. You review code, share patterns, and help colleagues improve their AI tooling workflow. You show what an AI-native approach looks like in practice — not by talking about it, but by doing it.

Technologies you will work with

• Python (FastAPI, asyncio) — Backend stack
• MCP (Model Context Protocol) — Tool integrations for AI agents
• Claude Code / Cursor / AI-native tooling — Your primary development environment
• PostgreSQL — Data storage and session management
• WebSockets — Real-time communication
• Docker — Containerized deployment
• LLM pipelines — Integration and orchestration of AI functionalities
• Event-driven architecture — Async processing and message-driven design

What you bring

Must-haves

• Your development workflow is AI-native. You use Claude Code, Cursor, or similar tools as your primary development environment — not as an occasional shortcut. You work with skills, plugins, and MCP servers to multiply your productivity. You can explain which tools you use, why, and how they change your workflow.
• You build backend systems in production. Several years of experience designing and running scalable services. You know what it takes to keep systems reliable under load — and you have proven this in practice.
• Python. Strong. Async. Production-ready. You know FastAPI and asyncio not from a tutorial, but from real systems.
• You think in systems, not in tickets. You see how services interact, where bottlenecks are, and how to design architecture that grows with the business. You make pragmatic choices — no over-engineering, no technical debt due to haste.
• You learn continuously. Tooling changes every month. What is state-of-the-art today is outdated tomorrow. You actively follow developments, experiment with new tools, and bring useful innovations back to the team.
• Fluent in English. Dutch is a plus.

Nice-to-haves

• Experience with MCP (Model Context Protocol) — building or consuming tool integrations
• Knowledge of LLM applications in production — how agents work, what they need from the backend
• Experience in regulated domains (finance, healthcare, legal)
• Background with event-driven architecture and message queues
• Experience in a startup environment or with lean teams

What you don't need to be

• An AI researcher. We have AI engineers who build the agents. You build the platform they run on.
• A DevOps engineer. We have infra people.
• A full-stack developer. We have frontend engineers.

Why Huub

• AI-native from day one. No 15 years of technical debt. No legacy systems to work around. We build greenfield with the latest tools, patterns, and models. This is not "we are migrating our monolith to microservices" — this is building as if it were 2026. Because it is.
• You work faster than you ever have. Not by working harder, but by systematically leveraging AI. Claude Code with custom skills and MCP integrations is not a nice-to-have — it's how we work. Engineers who embrace this build in days what used to take weeks.
• Small team, big impact. You are directly next to the decision-makers. No layers of management. Our modus operandi: build, ship, learn, repeat. Your architectural choices have a direct impact on the product and users.
• You build something that matters. We automate processes that are currently slow, error-prone, and frustrating. The systems you build make that possible — reliably, scalably, and securely.

Practical Information

• Location: Hybrid (Netherlands). Office in Amsterdam.
• Employment: Full-time
• Team: Engineering team of ~10 people
• Reports to: Mark Hinch
• Salary: Competitive, depending on experience
• Start: As soon as possible

Do you recognize this?

• "I had to build a new MCP server for an external data source. Instead of manually writing the boilerplate, I created a Claude Code skill that generates the server structure based on the API spec. The first version was ready in an hour. Now the entire team uses the same skill for new integrations."
• "We were getting timeouts on our agent pipeline at peak load. I added tracing, found the bottleneck in a sync database call, made it async, and added a circuit breaker. In production the same afternoon. The AI tooling made it easy to quickly generate and test the correct code — I only had to make the architectural decision."

If this sounds like a normal work week for you: let's talk.

How to apply

Send your CV and cover letter, tailored to this position, to vincent.lindeboom@huub.nl

If you (or someone you know) wants to be part of a motivated, lean team working on advanced technology, let's talk!

Key skills/competency

• AI-Native Development
• Python Backend
• FastAPI & Asyncio
• MCP Integration
• Event-Driven Architecture
• Scalable Systems
• Observability
• LLM Pipelines
• Docker
• PostgreSQL