---
title: 'GPU Cluster Engineer (Mensch) at Neura Robotics'
canonical: 'https://feeny.ai/job/gpu-cluster-engineer-mensch-neura-robotics-metzingen-06sgpmyef57f'
type: 'job'
last_seen: '2026-11-10'
---

# GPU Cluster Engineer (Mensch) at Neura Robotics

- **Company:** [Neura Robotics](https://feeny.ai/companies/neura-robotics)
- **Location:** Metzingen, Germany / Riederich, Germany
- **Employment:** full-time
- **Work type:** onsite
- **Posted:** 2026-04-13
- **Last confirmed live:** 2026-11-10
- **Apply:** https://jobs.ashbyhq.com/neura-robotics-gmbh/689ed8e5-c487-487d-963e-a9fc13483834/application

**Skills:** AWS HyperPod, Slurm, Kubernetes, Distributed Training, Cloud Cost Management, Infrastructure as Code, GPU Cluster Engineering, HPC Environments, AWS Instances, German Language

> The GPU Cluster Engineer serves as the central contact for NEURA's large-scale AWS HyperPod infrastructure, focusing on cluster engineering, operations, and stability. Responsibilities include designing self-service tools, optimizing GPU utilization, and collaborating with AWS to shape the platform at the hyperscale...

## Job description

Deine Mission & Herausforderungen

- Du bist die zentrale Ansprechperson für NEURAs GPU‑Cluster-Infrastruktur – ein großskaliges AWS‑HyperPod‑Setup mit topmodernen GPU‑Instanzen für Foundation‑Model‑Training und kundenspezifische Fine‑Tuning‑Workloads.
- Du entwickelst das Betriebsframework, baust Self‑Service‑Tools für die ML‑Teams und arbeitest direkt mit AWS zusammen, um die Plattform auf Hyperscaler‑Ebene mitzugestalten.
- Dein Fokus liegt voll auf Cluster Engineering & Operations — nicht auf ML‑Forschung selbst, sondern darauf, dass die Leute, die forschen, eine extrem stabile, effiziente und leicht zugängliche Infrastruktur haben.
- Aufsetzen, Konfigurieren und kontinuierliches Weiterentwickeln der HyperPod‑Cluster von NEURA, inkl. HyperPod/Slurm und HyperPod/EKS‑Orchestrierungsmodellen.
- Design und Umsetzung von Strategien für Cluster-Stabilität: Node‑Failure‑Detection, automatische Job‑Recovery, Checkpoint‑Koordination und fehlertolerante Multi‑Node‑Training‑Workflows.
- Aufbau eines Workload‑Priority‑Frameworks, das mehreren Teams und Use Cases – Pretraining, Fine‑Tuning, Kundenjobs – erlaubt, Clusterkapazität fair und effizient zu teilen.
- Optimierung der End‑to‑End‑GPU‑Auslastung: Erkennen und Lösen von Bottlenecks in Compute, GPU‑Speicher, EFA‑Netzwerk und Storage‑Durchsatz.
- Enge Zusammenarbeit mit den AWS HyperPod‑Produkt‑ und Engineering‑Teams: Issues eskalieren, Learnings aus einer der größten Deployments teilen und Anforderungen für die Roadmap platzieren.
- Bereitstellung von Self‑Service‑Tools, damit ML‑Researchers und Engineers Trainingsjobs eigenständig starten, monitoren und managen können – ohne ständige Infrastrukturunterstützung.
- Erstellung von Onboarding‑Dokus, Trainingsmaterial und internen Workshops, damit User effizient arbeiten, Best Practices einhalten und Kosten ihrer Workloads verstehen.
- Infrastructure as Code ist für dich Standard. Jede Cluster‑Konfiguration, jede Änderung, jede Umgebung ist Code‑first.
- Verantwortung für Kosten- und Kapazitätsstrategie: Spot‑Management, Reserved‑Instance‑Planung, Savings Plans und laufende AWS‑Commitment‑Verhandlungen.

Auf was können wir uns freuen

- 5+ Jahre Erfahrung im Infrastructure‑ oder Systems‑Engineering, idealerweise mit Fokus auf GPU‑Cluster oder HPC‑Umgebungen.
- Tiefe praktische Erfahrung mit AWS HyperPod und AWS‑Instanzen; direkte Erfahrung mit HyperPod ist ein starker Vorteil.
- Solides Verständnis von Slurm und Kubernetes als Orchestrierungsschichten – und die Fähigkeit, ihre Trade‑offs für große GPU‑Workloads zu bewerten.
- Praktisches Wissen über Distributed Training – du weißt, was Durchsatz beeinflusst und wie man Probleme debuggt.
- Erfahrung in der Entwicklung von Self‑Service‑Tools und technischer Dokumentation für anspruchsvolle Endnutzer: Du machst komplexe Infrastruktur zugänglich, nicht nur funktionsfähig.
- Starkes Verständnis für Cloud‑Kostenmanagement im großen Maßstab: Spot‑Interruptions, Kapazitätsreservierungen, Kostenverteilung über Teams und Workloads.
- Wohlfühlen in der Zusammenarbeit über Teamgrenzen hinweg – deine Hauptpartner sind ML‑Forschende, aber auch Product, Finance und Cloud‑Vendors.
- Sehr gute Englischkenntnisse; Deutsch ist ein Plus.

## About Neura Robotics

## Company Overview
- **One-liner**: NEURA Robotics develops cognitive robots that see, hear, feel, and learn, designed to work alongside humans in industrial, service, and household settings.
- **Entity Type**: Private (Series C — raised up to $1.4 billion as of 2025/2026)
- **Headquarters**: Metzingen, Germany (with offices in Bielefeld and Munich, and US locations in preparation)
- **Founded**: 2019
- **Founders**: David Reger

## Core Business
- **Primary industry/industries**: Cognitive Robotics, Humanoid Robotics, Industrial Automation, AI
- **Target customers**: B2B (industrial clients, integrators, strategic partners); also B2C via household robots like MiPA
- **Mission or purpose statement**: "We serve humanity by creating robots that give people back their time for what matters most. Our mission is to expand human potential through cognitive robotics that work safely alongside us."

## Products & Services
- **4NE1**: Production-ready humanoid robot designed for industrial workflows and everyday assistance, powered by proprietary AI (AURA). Built for mass production.
- **MAiRA (Multi-sensing Intelligent Robotic Assistant)**: The world's first cognitive robot arm — a collaborative robot that senses, responds, and adapts.
- **LARA (Lightweight Agile Robotic Assistant)**: Collaborative industrial robot for agile, lightweight automation tasks.
- **MiPA (My Intelligent Personal Assistant)**: Humanoid service and household robot for home assistance.
- **MAV (Multisensing Autonomous Vehicle)**: Mobile transport robot for logistics and material handling.
- **NEURA OmniSensor**: Proprietary sensor technology enabling robots to perceive their environment.
- **Neuraverse**: A global platform connecting robots, developers, and partners into a shared intelligence network where every deployment makes each robot smarter. Also serves as a marketplace for robotics applications.
- **NEURA Gyms**: Real-world training facilities where robots learn complex tasks in controlled environments to generate high-quality data for AI model validation.

## Market Standing
- **Valuation/Market Cap**: Not publicly disclosed (private company)
- **Key Metric**: Total investment >€100 million in 2023; Series C up to $1.4 billion (2025/2026); 2024 revenue in mid-double-digit million euros; 2025 target: triple-digit million euros
- **Notable Investors/Partners**: Lingotto Investment Management, BlueCrest Capital Management, InterAlpen Partners, Vsquared Ventures, HV Capital, Delta Electronics, C4 Ventures, Primepulse, L-Bank, Volvo Cars Tech Fund. Strategic partners include Qualcomm, Bosch, Schaeffler, NVIDIA, and AWS.
- **Growth Signals**: Over 1,400 employees from more than 45 countries (up from 300+ in 2024). Four of the world's ten largest robotics companies use NEURA's technology under their own brand. Deployed across Europe, US, China, and Japan. Europe's first production-ready humanoid robot (4NE1) launched.

## Competitive Advantages
- **Full vertical integration**: Develops all software, hardware, and platform technologies in-house, giving full control over performance, safety, and reliability.
- **Cognitive robotics pioneer**: Coined the term "cognitive robotics" — robots that perceive, learn, and adapt in real-time, unlike traditional industrial robots.
- **Neuraverse platform**: World's first platform connecting cognitive robots into a shared intelligence network, creating a data moat as every deployment improves all robots.
- **White-label partnerships**: Four of the top 10 global robotics companies OEM NEURA's technology, creating deep industry lock-in.
- **Diverse product range**: From collaborative robot arms (MAiRA, LARA) to humanoids (4NE1, MiPA) and autonomous vehicles (MAV), covering industrial, service, and household use cases.

## Strategic Focus
- **Scaling production**: Rapidly scaling robot production to meet growing demand, with Europe's first production-ready humanoid robot.
- **Expanding the Neuraverse**: Building the world's largest repository of real-world robotics data through shared learning across all deployed robots.
- **Global expansion**: Establishing US operations and strengthening presence in Japan and China.
- **B2B platform strategy**: Moving beyond hardware to become a platform company via the Neuraverse marketplace and NEURA Gyms.
- **Partnership ecosystem**: Deepening relationships with strategic partners like Bosch, Schaeffler, Qualcomm, NVIDIA, and AWS.

## Why Work Here
- **Culture**: "Passion for Winning" — achievement-driven, self-responsible culture with flat hierarchies. Values include trust, honesty, speed, and "We Are Human" (people at the heart of everything).
- **Flexibility**: Flexible working hours, 30 days of vacation, and a hybrid-friendly environment.
- **Team**: Over 1,400 employees from more than 45 countries — highly skilled, international team of experts.
- **Growth & Development**: Support for personal and professional growth via Udemy training platform and more.
- **Compensation**: Competitive salary package with exclusive employee discounts.
- **Perks**: Summer parties, company town hall meetings, and regular team celebrations.
- **Impact**: Opportunity to work on cutting-edge cognitive robotics and AI that directly addresses labor shortages and makes a meaningful societal impact.
- **Diversity**: Explicitly welcomes applications from all backgrounds, emphasizing building a team that reflects the world.
- **Application Process**: Fast — initial feedback within a few days, interviews scheduled within two weeks.

## Sources
1. [neura-robotics.com](https://neura-robotics.com/)
2. [press.neura-robotics.com](https://press.neura-robotics.com/factsheet-neura-robotics-en/)
3. [neura-robotics.com/company-about-us](https://neura-robotics.com/company-about-us/)
4. [jobs.neura-robotics.com](https://jobs.neura-robotics.com/)
5. [neura-robotics.com/press](https://neura-robotics.com/press/)

## Other roles at Neura Robotics

- [Linux Infrastructure Engineer - NeuraGym (Mensch)](https://feeny.ai/job/linux-infrastructure-engineer-neuragym-mensch-neura-robotics-stuttgart-6pwpmy7ncx7a) — Stuttgart, Germany
- [Linux Infrastructure Engineer - NeuraGym (human)](https://feeny.ai/job/linux-infrastructure-engineer-neuragym-human-neura-robotics-stuttgart-hnjeef65epw3) — Stuttgart, Germany
- [Network & Systems Administrator – RoboGym (Human)](https://feeny.ai/job/network-systems-administrator-robogym-human-neura-robotics-stuttgart-t8xaq61a1d1g) — Stuttgart, Germany
- [Network & Systems Administrator – RoboGym (Mensch)](https://feeny.ai/job/network-systems-administrator-robogym-mensch-neura-robotics-stuttgart-yx3pzja8xe11) — Stuttgart, Germany
- [Network & Systems Administrator – RoboGym (Mensch)](https://feeny.ai/job/network-systems-administrator-robogym-mensch-neura-robotics-munich-r12t6nje4nws) — Munich, Germany / Freising, Germany
- [Network & Systems Administrator – RoboGym (Human)](https://feeny.ai/job/network-systems-administrator-robogym-human-neura-robotics-munich-j41yg4cpt166) — Munich, Germany / Freising, Germany
- [Programme Manager – NEURA Research Programme (human)](https://feeny.ai/job/programme-manager-neura-research-programme-human-neura-robotics-metzingen-fqesea0mc426) — Metzingen, Germany / Riederich, Germany
- [Customer Project Manager (mensch)](https://feeny.ai/job/customer-project-manager-mensch-neura-robotics-munich-wvgm5ry2c9xf) — Munich, Germany
- [Customer Project Manager (human)](https://feeny.ai/job/customer-project-manager-human-neura-robotics-munich-e28fzskp9aj8) — Munich, Germany
- [Customs & Trade Compliance Manager (m/f/d)](https://feeny.ai/job/customs-trade-compliance-manager-m-f-d-neura-robotics-zurich-3w1angvh78rp) — Zurich, Switzerland