The One- Line Vision:
We're looking for engineers who have taken AI systems from working demos to production systems that real users depend on.
If you've only wired prompts together and called it AI engineering, this isn't your role.
If you've fought hallucinations in production, built evaluation frameworks, debugged agent failures, and care deeply about reliability, cost, and customer impact, keep reading.

Why Max:

MAX is Beroe's AI-native, always-on decision engine that makes procurement continuously competitive.It is the missing connecting layer between intelligence and execution, combining millions of live market signals, customer spend, contract and supplier data, Beroe's market intelligence, and codified procurement methodologies.MAX continuously senses, reasons, and surfaces the decisions that need attention—moving procurement from reacting to anticipating, episodic to continuous, and information overload to decision intelligence.

Why Now:
The platform is already live and in the hands of customers. The next generation of capabilities is largely unbuilt, and you will help define it.

What you will build:

Production-grade multi-agent systems.
Large-scale Retrieval-Augmented Generation (RAG) systems.
AI memory, evaluation, and observability frameworks.
Recommendation engines and autonomous monitoring agents.
Explainable decision intelligence and enterprise knowledge systems.

Problems you will tackle:

Multi-agent orchestration and long-running workflows
Multi-agent orchestration and long-running workflows
AI memory, evaluation, and observability frameworks.
AI memory, evaluation, and observability frameworks.
Explainable decision intelligence and enterprise knowledge systems.

Responsibilities

Own platform capabilities end-to-end.
Design resilient distributed systems.
Build scalable backend services and APIs.
Improve reliability, performance, and cost efficiency.
Raise engineering standards and mentor other engineers.

Join MAX — Beroe's AI-native, always-on decision engine.

What are we looking for:

7+ years of software engineering experience.
Strong software engineering fundamentals and deep proficiency in Python.
Experience building and operating AI systems in production.
Experience with LLM applications, RAG, tool calling, and multi-step workflows.
Experience evaluating AI quality and strong debugging skills.

Strong Plus:

Modern AI orchestration frameworks (LangGraph, LlamaIndex, or similar).
Vector databases and knowledge graphs.
AI observability platforms.
Large enterprise datasets and event-driven architectures.

Hiring Philosophy

We care more about systems you've built than frameworks you've used.
We value ownership, pragmatism, and engineering excellence.

The Role is probably Not for you if:

Your experience is primarily demos and hackathons.
You've never operated systems in production.
You over-index on prompt engineering.
You prefer experimentation over ownership and delivery.

At Beroe X nnamu GmbH, we prioritize a balanced work-life experience.

Here’s what we offer you:

4-Day Work Week & up to 24 Days Paid Vacation – More time to recharge
Competitive Compensation – Fair salary and comprehensive benefits.
Flexible Work Arrangements – Work the way that suits you. Professional Development – Access to training and certifications.
Team Events – Bi-annual company events and quarterly lunch get-togethers.
Work Abroad Flexibility – Remote work from the EU or selected non-EU countries for up to 8 weeks a year with travel insurance coverage

We operate on a hybrid model with offices in Berlin and Munich, offering a 32-hour, 4-day workweek.
This means:

In-Office Collaboration: Work from the office two days a week
Manage Your Own Hours: Flexibility to work around your needs as long as team goals are met.

Our Culture  
We are committed to fostering a collaborative, innovative and inclusive work environment where everyone’s ideas matter. We know that diverse teams lead to better outcomes and welcome applicants from all backgrounds.

Senior Software Engineer – Agentic Systems (m/f/x)