Please note:
On this page you will only see the English-language presentations of the conference. You can find all conference sessions, including the German speaking ones, here.
The times given in the conference program correspond to Central European Time (CET).
Leveraging AI as an Intelligent Assistant for 24/7 Incident Resolution
This talk presents an experience report on utilising Artificial Intelligence (AI) as an intelligent assistant for a team of DevOps Engineers responsible for operating a critical customer system 24/7. Given the complexity and criticality of the supported system, the DevOps team has built a comprehensive library of how-to guides, documenting specific incident cases and their solutions.
However, accessing and following the appropriate guidelines requires diligent analysis, experience, and a high level of concentration from the on-duty personnel. To address this challenge, we have developed an AI-based assistant using a Large Language Model (LLM), which takes the incident description as input and guides the DevOps Engineer through the necessary steps to resolve the incident.
The AI assistant performs the following functions:
- Suggests the appropriate incident resolution process by matching the error description with the how-to library.
- Parameterises necessary database or log queries specific to the incident case.
- Generates intermediate steps required for documentation purposes.
During this talk, we will present the system setup, the AI-based services employed, and the challenges encountered in providing context information to the LLM. Additionally, we will discuss our learnings and solutions regarding the selection of suitable prompts for the LLM.
The insights shared in this experience report shed light on the practical implementation of AI as an intelligent assistant for 24/7 incident resolution. By leveraging AI technologies, organisations can enhance their incident resolution processes, improve efficiency, and reduce the reliance on human expertise.
Target Audience: Developers, Project Leader, Decision Makers
Prerequisites: None
Level: Basic
Tim Chen is a Software Engineer at MaibornWolff, specializing in AI-driven solutions. He has contributed to the development of AI-assisted tools, leveraging his expertise in natural language processing and large language models. With a strong foundation in software engineering, Tim has significant experience in full-stack development and AI integration through his involvement in various projects.
Benjamin Hofmann has a rich background in Software Testing & Engineering, particularly in Cloud Architectures and recently GenAI Solutions. Throughout his career, he has advanced evolving fields such as Mircoservices, Platform Engineering and Observability while implementing Agile and DevOps methodologies. As a founder of the AI4U Initiative at MaibornWolff, he is dedicated to advancing the effective use of generative AI within the company and its software projects.
Vortrag Teilen
