Internship Software Engineering: Python & AI Agents (Business Process Automation)

Description du poste

Join MESTRO to develop the next lifecycle step of ELCAs AI-driven BPM orchestrator. Tasks Transition approved apps into real Flowable app packages. Build core translation between MESTRO and BPMS platforms. Incorporate independent test cases into Flowable's testing. Skills Degree in computer science or related fields required. Strong knowledge of Python and software engineering fundamentals. Experience with LLMs and agent-based AI preferred. This job ad was AI-translated from German. See original About the job You will help build the next lifecycle step of MESTRO after analysis and design have already been completed. MESTRO is ELCA's AI-powered BPM orchestrator: it manages business processes from the initial idea to a running, automatically tested system. Your task is to build the bridge to real BPMS execution. AI should not only generate code but actively detect and minimise which information is lost when translating process models and how this can be prevented. You will learn how to orchestrate AI agents production-ready, not as a demo but as a feature that customers actually use. Your tasks You will bring MESTRO through the instrumentation and testing phase, which is divided into three work packages. The roundtrip core is your main focus: App handover from design (WP1): You transfer approved apps from the design phase into a real Flowable app package, automate deployment via the REST interface, and extend the existing MCP server with the tools defined in the architecture concept. Roundtrip between MESTRO and BPMS platforms (WP2 & main focus): You build the core that translates models visibly lossless to Flowable and returns platform additions via AI-supported diff & merge. Overwriting never happens automatically, as every change is actively confirmed. You control the edit-deploy-sync cycle together with Claude code agents via MCP tools. Transfer test cases into Flowable (WP3): You bring MESTRO's tool-independent test cases into Flowable's in-product testing as a basis for the upcoming test & fix phase. You actively shape the prioritisation within the packages. Delivery is monthly in the form of visible, functioning increments. Our offer A dynamic environment with a motivated team at the interface of business and IT, where you create real impact with modern workflows and AI automation Access to innovative technologies such as low-code platforms, AI tools, RAG and agentic coding (e.g. Claude Code and Cowork) A flat hierarchy and a culture of collaboration across all disciplines The chance to change people's lives by building great software Work-life balance, ergonomic and flexible workplace in a multicultural environment Your profile Completed degree (university, ETH, FH) in (business) computer science or related disciplines Solid knowledge of Python (FastAPI) and a solid software engineering foundation (REST APIs, Git, Docker, testing) Strong interest and ideally practical experience with LLMs and agent-based AI (prompt engineering, agent frameworks, model context protocol) Structured, independent working style and analytical thinking Experience in frontend (Vue.js), CMMN, experience with a BPMS/workflow engine (Flowable, Camunda or similar) is an advantage Fluent knowledge of German and English, spoken and written We rely on healthy values and are convinced that diversity is the key to success. Those who approach challenges with curiosity, motivation and drive are exactly right with us! Be inspired by our motto: We make it work!