Senior Full Stack Engineer, Agentic Workflows (m/f/d)

RobCo; Deutschland; Remote; EUR 90,000 - 150,000; Full time

Benefits offered by this job

Wellpass membership

Flexible working hours

Office in Munich

Work from home

Job summary

RobCo is developing autonomous industrial robotics with a full AI stack—hardware, software and AI in one platform. We seek a senior engineer to design, build, and operate agentic workflows across delivery, ensuring scalable, secure and observable systems.

You will ship LLM-powered solutions, craft prompts, and integrate with team tools to empower non-engineering colleagues, while maintaining robust architecture and strong testing discipline. Munich-based hybrid flexibility offered.

Qualifications

• Shipped LLM-powered or agentic systems that real users depend on.
• Strong context engineering and prompt engineering skills; prompts treated as versioned engineering artifacts.
• Knowledge bases and RAG pipelines for reliable information access.
• Systems thinking: design for multiple agents and scalable end-to-end solutions.
• Comfort working with non-engineering teams to automate real workflows.

Responsibilities

• Design, build, and run agentic workflows across our delivery process—from requirements review to production.
• Build reusable components, tools, and patterns for integrations, memory, and secret management.
• Own end-to-end quality: evals, observability, cost, and guardrails for trustworthy agent outputs.
• Design human-in-the-loop workflows to learn from corrections and improve prompts and components.
• Embed with delivery teams to maximize leverage of agent capabilities.

Skills

Software engineering

Python

TypeScript

LLM/agentic systems

Prompt engineering

System design

Tools

MCP framework

Job description

About the Role at RobCo

RobCo is building the category of Autonomous Industrial Robotics and we're doing it fast. We believe that automating the ordinary frees humans to do the extraordinary, and we're building the end-to-end robotics infrastructure to make that possible - hardware, software, and AI in one stack that just works.

Your Responsibilities

• Design, build, and run agentic workflows across our delivery process - requirements review, automated checks and gates, early risk detection, and a knowledge base that makes everything we've learned from past projects available to agents and people - owning each from concept through production.
• Build the way we build: reusable components, skills, and patterns for tool integrations, feedback loops, and identity, secret, and auth management. You know the agentic toolbox inside out and decide what we actually need and what we don't.
• Own quality end to end: design and run evals, monitor agents in production, and manage observability, cost, guardrails, and security so teams can trust what the agents produce.
• Design human-in-the-loop workflows that learn: people review and correct agent output at defined checkpoints, and those corrections flow back into better prompts, evals, and components, so the agents keep improving.
• Embed with our delivery teams to understand how their work actually flows, identify where agents create the most leverage, and define together what good looks like.
• Drive adoption and help shape how RobCo uses AI internally: deliver agents through the tools teams already use every day, establish best practices, and build paved paths so teams can eventually create their own agents.

We're looking for someone who has:

• Several years of software engineering experience and the seniority to take a problem from ambiguous idea to production system, with clean architecture, solid testing, and maintainable code.
• Shipped LLM-powered or agentic systems that real users depend on, not just prototypes: tool calling, structured outputs, multi-step workflows. Deep, practical experience with the agentic stack — orchestration, tool use, MCP, memory, and retrieval — in production environments.
• Strong context engineering and prompt engineering skills: you think carefully about what the model sees and in what structure, you treat prompts as versioned, tested engineering artifacts, and you have a rigorous way of judging whether the output is actually good.
• Built knowledge bases and put RAG pipelines to work. You don't need to be a retrieval specialist; you need to make organizational knowledge reliably available to agents and people.
• Thinks in terms of systems and user outcomes, not individual scripts and demos, and brings a component mindset: while building the first agent, you're already thinking about the second and the tenth.
• Genuine enjoyment of working with non-engineering teams: understanding their process, speaking their language, and designing automation around how they actually work.
• Strong proficiency in Python and/or TypeScript, plus fluency with SaaS APIs, webhooks, and the practical realities of integrating business systems.
• Ownership of outcomes and a startup way of working: scope intentionally, build smart, ship fast, and prefer pragmatic solutions that ship over perfect ones that don't.
• Fluent English (German is a plus).

Why join RobCo?

• Develop modular, software-defined robotics solutions with us that set new standards in automation.
• Help us build robots equipped with state-of-the-art AI that perceive their environment, analyze data, and make intelligent decisions in real time.
• As part of a very dynamic and fast-growing scale-up, you will take on responsibility from day one, drive your own projects, and see the direct impact of your work.
• Flat hierarchies, open communication, and regular 360° feedback enable fast decisions and support your professional and personal development.
• Wellpass membership (7,200+ studios across Germany), regular team events, and a team that takes both the work and the culture seriously.
• Flexible working hours, a modern office in Munich whenever you with to visit, and the flexibility to work from home.
• Regular team events - from relaxed beer & pizza nights on the rooftop terrace to shared …