AI / Software Engineering Werkstudent (m/f/d) — Coding Agents & Developer Tooling

About

We’re looking for a curious and technical student to explore the future of coding agents and developer tooling.

This is not a typical “implement tickets” internship.

The role is intentionally research-oriented and experimental: we want someone who enjoys thinking deeply about how AI coding agents work, where they fail, and how better tooling/interfaces can improve their effectiveness.

What you’ll work on

You’ll help research and prototype:

• AI coding agents
• autonomous developer workflows
• CLI tooling for LLM agents
• agent orchestration
• developer productivity systems
• codebase navigation and context management
• prompt / memory / tool abstractions
• evaluation and benchmarking setups

The exact scope is intentionally flexible. A big part of the role is brainstorming ideas together, rapidly testing concepts, and building small experimental tools.

Who we’re looking for

You are probably:

• currently pursuing a Master’s degree in Computer Science, AI, Software Engineering, Robotics, Mathematics, or a related technical field OR
• an exceptional self-taught builder with strong GitHub projects and evidence of technical curiosity

We care much more about:

• technical curiosity,
• ability to build,
• independent thinking,
• and genuine interest in AI systems

than formal credentials.

Strong signals

• personal projects
• open-source contributions
• CLI tools
• AI agent experiments
• LLM tooling
• developer tooling experience
• strong GitHub profile
• hacker / builder mindset

Tech stack (likely)

• Python
• TypeScript / Rust
• terminal / CLI tooling packages
• LLM APIs
• local inference tools
• agent frameworks
• developer tooling ecosystems

What this role is like

This is closer to:

• applied research,
• hacking,
• prototyping,
• and technical exploration

than a traditional internship.

You’ll have significant freedom and ownership, and will work directly on early-stage ideas around AI-assisted software engineering workflows.

Details

• Location: Berlin or remote within Germany
• Work can be done remotely BUT we will need to be able to regurlarly meet, therefore, candidates based in Berlin are preferred
• Type: Werkstudent / part-time
• Hours: flexible (typically 10–20h/week)
• Language: English-friendly environment
• Compensation: competitive depending on experience and profile
• Start date: flexible

How to apply

Please send:

• your GitHub profile
• LinkedIn or CV
• a few projects you’ve built
• and a short note about why this topic interests you

We are especially interested in candidates who spend their free time experimenting with AI tooling, agents, terminals, developer workflows, or open-source projects.

Applicants of all backgrounds and identities are welcome.