Allgemeine Informationen
| Veranstaltungsname | Übung: 5487UE AI-Driven Software Development |
| Untertitel | |
| Veranstaltungsnummer | 5487UE |
| Semester | SoSe 25 |
| Aktuelle Anzahl der Teilnehmenden | 95 |
| erwartete Teilnehmendenanzahl | 60 |
| Heimat-Einrichtung | Lehrstuhl für Software Engineering II |
| Veranstaltungstyp | Übung in der Kategorie Lehre (mit Prüfung) |
| Erster Termin | Dienstag, 29.04.2025 10:00 - 12:00 Uhr, Ort: (IM) R 028 |
| Art/Form | |
| Voraussetzungen |
Empfohlen: Programmierung I + II, Software Engineering, Software Testing |
| Lernorganisation |
The course covers the following topics: • Prompt Engineering • AI Agents for Software Development • Requirements Engineering • Prototyping • AI-Assisted Coding • AI-Driven Testing and Quality Assurance • Refactoring and Code Cleaning • Development Processes and Collaboration |
| Leistungsnachweis |
Portfolio-Prüfung. Die genauen Anforderungen werden vom Dozierenden zu Beginn der Veranstaltung bekanntgegeben. Mögliche Portfoliobestandteile umfassen: • Dokumentierter und ausführbarer Quellcode der Projektaufgaben • Live-Demonstration der implementierten Softwarelösung • Technischer Bericht über die Nutzung und Erfahrungen mit dem KI-gestützten Prozess • Laufende technische Teilberichte, zusammengefasst in einem abschließenden Gesamtbericht • Präsentation der Arbeit nach jedem Teilabschnitt des Projekts • Verlauf der verwendeten Prompts • Versionshistorie, z. B. mit Git |
| SWS |
2 |
| Qualifikationsziele |
Kenntnisse / Skills / Knowledge Students deepen their knowledge of software engineering by integrating modern AI-driven tools and methods into the practical development of software projects. The focus is on applying LLMs to support classical software engineering tasks such as requirements analysis, prototyping, coding, test automation, and refactoring. Additionally, students learn various prompting strategies and how to differentiate between classical information retrieval techniques and AI-powered approaches. Fähigkeiten / Abilities Participants develop the ability to effectively apply AI-driven software development by integrating LLMs into existing development processes. They can leverage AI-powered tools to automate software testing, improve code quality, and enhance development efficiency. Additionally, they gain an understanding of how to adapt LLMs to different project requirements and design appropriate interfaces and workflows. Kompetenzen / Competencies Participants acquire hands-on competencies in developing and implementing projects using AI-driven software solutions. They are able to combine modern software engineering techniques with LLMs to address challenges in areas such as requirements management, code generation, quality assurance, and automated testing. Additionally, they learn how to strategically integrate AI-driven development approaches into larger software projects, including evaluating cost-benefit aspects and long-term maintainability. |
| Workload |
30 contact hours + 120 hours independent study Exercise: Introduction to common software engineering tasks and how AI can support them. Explanation of the projects and requirements for portfolio components as well as questions about the task and the respective solution approaches Independent study: Processing of projects |
| ECTS-Punkte |
5 |
