Software Engineer (AI)

Overview

One of Munich’s strongest early-stage AI startups, already live with multiple high-value customers and embedded in real industrial workflows. They’re building AI systems that sit directly inside complex manufacturing sales processes, not just dashboards or copilots.

The product is already in market with paying customers and a small number of deep design partners. Backed by tier-1 investors and experienced founders, the team is still early and highly selective. Strong technical bar, strong commercial pull.

What makes this interesting is the depth of the problem. You’re dealing with messy, multimodal data and real decision-making systems that impact how machines get configured and sold. This is not experimentation, it is production from day one.

Role

Founding-level AI Engineer building and shipping agent-based systems into live customer environments.

Responsibilities

• Design and build AI agents across extraction, reasoning and configuration workflows
• Own end-to-end systems from data ingestion through to deployment
• Work directly with customers to understand real-world constraints and translate into working systems
• Build and scale backend services for AI workflows in production
• Contribute to architecture decisions across agents, infra and data pipelines
• Continuously improve performance, reliability and evaluation of AI systems

Requirements

• Strong software engineering fundamentals with Python
• Proven experience building and deploying LLM or agent-based systems
• Comfortable working across backend systems and data pipelines
• Ability to operate in a fast-moving, low-structure environment
• Strong problem-solving mindset, able to deal with ambiguity and messy data
• Interest in working close to customers and real-world systems

Python, Temporal, OpenAI / Anthropic / Gemini, LiteLLM, Langfuse, Pydantic AI

Postgres, vector and graph databases

Next.js, TypeScript

Requirements / constraints

Munich-based, in-office environment

Package

They will pay what it takes within reason for the right person. Compensation is flexible and tied to impact, not bands.