Skip to content

The Automation Hour - 10

AutoThumb10

  • KLPT
  • MBRH

(1:52) Thomas introduced the meeting, mentioning that Simon was attending a training course on Ansible. He outlined the agenda, which included two main topics: PS Net Know it All by Kenneth and Martin, and ways to create knowledge articles by Karthick.

(4:17) Martin provided an overview of PS Net Know it All, explaining its purpose in utilizing AI to resolve incidents quickly. He highlighted its use in identifying production-critical issues and assigning incidents to the appropriate teams.

(9:41) Martin demonstrated the features of PS Net Know it All, including the ability to view incidents, alerts, and knowledge base articles. He emphasized the importance of adding detailed incident resolutions to improve the AI’s effectiveness.

(16:10) Kenneth explained the architecture of PS Net Know it All, detailing the virtual machine setup, data sources, and integration with ServiceNow. He also discussed the planned transition to a different LLM model and the use of Azure AI Foundry.

(27:40) Kenneth described the functionality of PS Net Know it All, including how it processes alerts and incidents using keywords and prompts. He explained the steps involved in generating responses and the importance of keywords in linking queries to relevant data.

(42:44) Karthick presented the process of creating knowledge articles, emphasizing the importance of well-crafted articles for quick problem resolution. He demonstrated how to use templates and transform closed incidents into knowledge articles.

(43:45) Karthick introduced the PA KB candidate tag, which allows engineers to mark incidents that can be transformed into knowledge articles. This helps in identifying potential articles and streamlining the creation process.

(51:38) Karthick mentioned the process of reviewing and maintaining knowledge articles, including periodic reviews by the QC team to ensure the articles remain up-to-date and relevant.