Powering Generative AI with Semantic Kernel
summary
A runnable pipeline ingests Sentry error logs and documentation (including Cognity Search) into Semantic Kernel connected to OpenAI, then auto-generates either shell commands (e.g., restarting Kubernetes pods) or code edits pushed as generated pull requests for self-healing. The demo shows an automated NullPointer fix created as a PR and a cluster restart executed with post-checks and notifications; it’s implemented as an Azure-oriented Kubernetes operator prototype that accelerates response but still requires human review to avoid LLM hallucinations.
AI-written summary · about
chapters (13)
- 0:00Digital immune systems
- 1:02Immune system goals
- 3:32Gartner's robust prerequisites
- 6:24NTT Data FinOps
- 7:28Self-healing with GenAI
- 9:59Self-healing vs auto-remediation
- 11:44Self-healing in Kubernetes
- 13:32Semantic Kernel benefits
- 16:45Solution architecture demo
- 18:50Error generation & detection
- 21:21Automatic analysis & correction
- 23:11AI-generated Pull Requests
- 25:17Q&A
related talks
26mSemantic Kernel: Teaching AI to Not Just Repeat, But Complete!
33mHow to enhance your own AI services using Semantic Kernel
01h 12mBuilding Scalable GenAI Apps Using Semantic Kernel and .NET Aspire
46mBuilding AI solutions with Semantic Kernel
01h 02mDevelop AI agents with Semantic Kernel
01h 00mBuilding your own AI Agent using Semantic Kernel
47mAI Orchestration with Semantic Kernel
32mKeynote: Powering Generative AI with Kubernetes