RAGnaroX — Secure Local ChatOps with Small Language Models
RAGnaroX presents a secure, locally hosted ChatOps assistant using small language models, hybrid retrieval, and function calling for efficient on-premise AI support.
I research and engineer sustainable software systems — from efficient web applications to resource-aware AIware — that reduce costs without compromising quality.
RAGnaroX presents a secure, locally hosted ChatOps assistant using small language models, hybrid retrieval, and function calling for efficient on-premise AI support.
My talk focused on how AI supports everyday motorsport club operations, balancing real-world benefits with essential guidance on data privacy, reliability, and transparency
The paper "Retrieval-Augmented Generation in Industry: An Interview Study on Use Cases, Requirements, Challenges, and Evaluation" received an Honorable Mention for the Best Paper Award at KDIR 2025.
Together with Maximilian Ehrhart, I hosted the GENIUS plenary meeting in Tyrol, bringing together international partners to discuss AI research advances, industry collaboration, and future directions in generative AI and software engineering.
A GENIUS collaboration with CASABLANCA hotelsoftware, the University of Innsbruck, and c.c.com Moser GmbH showed that smaller language models can deliver comparable results in RAG-based information retrieval, while reducing costs by around 75%.
The AIware 2025 paper presents a combined vision of industry and academia for the future use of generative AI across the software development life cycle.
arXiv:cs.SE
Transforming Privacy Artifacts into Accessible Reports for Non-Technical Stakeholders
arXiv:cs.SE
RESTestBench: A Benchmark for Evaluating the Effectiveness of LLM-Generated REST API Test Cases from NL Requirements
arXiv:cs.AR
RAGnaroX: A Secure, Local-Hosted ChatOps Assistant Using Small Language Models
arXiv:cs.IR
RAG-DIVE: A Dynamic Approach for Multi-Turn Dialogue Evaluation in Retrieval-Augmented Generation
arXiv:cs.SE
RESTifAI: LLM-Based Workflow for Reusable REST API Testing
2025 2nd IEEE/ACM International Conference on AI-powered Software (AIware)
The Future of Generative AI in Software Engineering: A Vision From Industry and Academia in the European Genius Project
Advances in Intelligent Systems and Computing
SmartDelta Methodology: Automated Quality Assurance and Optimization for Incremental System Engineering
Proceedings of the 17th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management
Retrieval-Augmented Generation in Industry: An Interview Study on Use Cases, Requirements, Challenges, and Evaluation
2024 50th Euromicro Conference on Software Engineering and Advanced Applications (SEAA)
How Industry Tackles Anomalies during Runtime: Approaches and Key Monitoring Parameters
Lecture Notes in Computer Science
Web Image Formats: Assessment of Their Real-World-Usage and Performance Across Popular Web Browsers
Lecture Notes in Computer Science
On the Role of Font Formats in Building Efficient Web Applications
Microprocessors and Microsystems
SmartDelta project: Automated quality assurance and optimization across product versions and variants
2023 IEEE/ACM 10th International Conference on Mobile Software Engineering and Systems (MOBILESoft)
Energy-Saving Strategies for Mobile Web Apps and their Measurement: Results from a Decade of Research
arXiv:cs.SE
SoHist: A Tool for Managing Technical Debt through Retro Perspective Code Analysis
SmartDelta was an international ITEA project advancing automated quality assurance for incremental software engineering. It developed tool-supported mechanisms to assess product deltas in continuous engineering environments, combining artifacts and runtime data to detect quality improvements and risks. Consortium: 34 partners across 9 countries including Alstom, Software AG, Akkodis, Kuveyt Türk Bank, and Vaadin.
GENIUS is an international ITEA project shaping the next generation of generative-AI-enabled software engineering. It investigates how LLMs and AI engineering assistants can be integrated across the SDLC — from requirements and code generation to testing, documentation, and quality analysis. Consortium: 32 partners across 9 countries including c.c.com, CASABLANCA hotelsoftware, British Telecom, and Siemens.
KI2Industrie is an FFG-funded qualification initiative strengthening the practical AI capabilities of companies in Western Austria. It establishes a modular continuing-education programme covering AI transformation, ML, LLMs, the EU AI Act, and industrial AI transfer. Consortium partners include Pfeifer Holding GmbH, VAHLE AUTOMATION GmbH, and TINETZ-Tiroler Netze GmbH.
benedikt.dornauer@outlook.at
Connect with me
View my projects
View my research