Want to know if this job is worth applying to?
München, Bayern, Deutschland
Onsite
SE3 Labs builds Spatial AI by combining 3D computer vision and large language models (LLMs). Ultimately, our AI will be able to answer any question about 3D data that a human analyst could answer. Spatial AI will unlock tremendous value in industries like construction, smart infrastructure, smart factories, and defense. As our long-term vision, our spatial AI is a crucial building block for enabling LLM-controlled humanoid robots.
Our team consists of AI experts and repeat founders, who have worked with Elon Musk at Tesla and at Nvidia, KPMG, Bosch, and Artisense.
Within our first year, we have secured multi-million dollar revenue and aim for mid eight-digits in 2026.
We are a dual-use tech company selling to civil customers and the German armed forces. We will expand to the UK and the Ukraine in the next three months.
SE3 Labs is building Spatial AI: we combine 3D computer vision with large language models to give drones the ability to understand and act in the physical world, then deploy that intelligence on real UAVs.
This internship exists for one reason: to ship world-class perception + AI capability into production systems, fast.
You’ll work closely with our founding engineers on projects that are materially important to the company. The work is hands-on, technical, and execution-driven, more like a top-tier autonomy team than a traditional internship.
Take messy, ambiguous problems and turn them into shipped product.
Own the full path from idea → implementation → rollout → feedback → improvement.
Turn LLMs into doers (Tools + MCP + RAG)
Build tool-using LLM systems that take real actions across DBs, services, and internal APIs to gather valuable insights.
Build MCP servers/clients and tool interfaces with real permissions and safe execution.
Implement RAG that holds up when it counts: ingestion, indexing, retrieval, reranking, evals, freshness, grounding.
Make agents genuinely reliable: retries, timeouts, fallbacks, guardrails, observability.
Work across the stack for your own module
Be effective anywhere the problem lives: pub/sub + eventing, backend services, APIs, Postgres/data, retrieval/indexes, and UI integrations.
Build with strong software engineering fundamentals: clean code, testing, debugging, performance, operability.
Design pragmatic AI architecture: the right mix of LLMs, tools, retrieval, caching, and workflows.
A problem-solver who takes ownership, unblocks themselves, and ships end-to-end.
Strong software engineering: clean, maintainable code and architecture; sharp debugging; performance-aware.
A serious LLM systems builder: tool calling + agents, RAG, MCP (not all required, but real competence in the area)
Backend fundamentals: APIs/services, Postgres, pipelines, observability.
A production mindset: you care about latency, data quality, and "does it actually work in real life?"
21 days of paid leave: We offer 20 days of annual leave, plus an extra day off to celebrate your birthday.
Team activities: We organize at least 2x company offsite per year. The last one was at ECCV in Milan and the next one will be in Greece.
Working long and intense hours requires proper fuel - that's why dinner is always on the company.
Fitness and Well-Being: Through our partner E-GYM Wellpass, we offer you access to a comprehensive fitness program - including premium gyms, climbing, and much more.
Intro Call (virtual, 30 minutes): This is our chance to get to know each other. We’ll discuss your background, your motivations, and why you’re interested in SE3 Labs. You’ll also learn more about our mission, the role, and the team.
LeetCode Interview (flexible, ~60 minutes): This will help us understand your approach to problem-solving and coding.
Dinner/lunch (in person): We would love to get to know you in person and tell you about how we work at SE3.
