Working Student - AI Platform and Enablement (m/f/d)

<p><strong>Mission Brief&nbsp;</strong></p> <p>Isar Aerospace is building the launch company Europe needs: fast, vertically integrated, technically ambitious, and execution-driven.</p> <p>We are currently building up our internal AI stack, including a self-managed inference endpoint for deploying open-source models on virtual machines. We are looking for a technically curious working student who can help operate and continuously improve this setup while staying close to the newest models, tools, and agentic best practices.</p> <p>This is a hands-on role for someone who enjoys making AI systems work reliably, exploring new capabilities, and identifying practical opportunities to automate engineering and business workflows.</p> <p>&nbsp;</p> <p>Your Role in Our Space Mission:&nbsp;</p> <p>Help keep Isar Aerospace's internal AI stack current, reliable, and useful, while exploring tools and workflows that make modern AI accessible to teams across the company.</p> <p><strong>Support with operation and maintaining the AI Stack</strong></p> <ul> <li>Support with the maintenance of the self-managed inference environment used to deploy open-source AI models.</li> <li>Update and test deployed models, Docker images, and supporting services.</li> <li>Configure and maintain the model gateway, including model routing and endpoint settings.</li> <li>Support integrations through REST APIs, MCP servers, and related interfaces.</li> </ul> <p><strong>Evaluate models and tools</strong></p> <ul> <li>Track relevant developments in open-source models, inference tooling, and agentic workflows.</li> <li>Run practical evaluations and recommend changes based on quality, usability, reliability, and resource requirements.</li> <li>Help maintain concise documentation so the stack can be operated and used consistently.</li> </ul> <p><strong>Client tools and automation (optional)&nbsp;</strong></p> <ul> <li>Curate and test useful client tools that work with the internal endpoint, such as coding harnesses, developer tools, and automation platforms including n8n.</li> <li>Explore the Microsoft agent platform, including Microsoft 365 Copilot agents and Copilot Studio, and build selected workflows or agents for internal use.</li> <li>Identify repetitive or information-heavy tasks that are worthwhile to automate, then prototype pragmatic solutions with relevant teams.</li> </ul> <p>&nbsp;</p> <p><strong>Qualification Checklist&nbsp;</strong></p> <ul> <li>Currently enrolled in computer science, data science, software engineering, information systems, or a related technical field.</li>