Supercharge Your PowerShell - Harnessing...
summary
Demos show local LLMs loaded via Olama + CodeGPT local models, a Microsoft AI toolkit hitting the NPU on a Snapdragon X Elite while CPU runs crawl, and a beefy RTX 4090 loading a 32B model; attempts sometimes crash the laptop but the NPU path works reliably. GitHub Copilot integrates with local models to review code and generate Pester unit tests and LLMs (Gemini 2.5, GPT‑4o) analyze 41‑file repos and suggest prioritized performance fixes.
AI-written summary · about
chapters (13)
- 0:00PowerShell and LLMs intro
- 2:06Learning objectives
- 4:31Local LLM setup with Olama
- 7:17Demo: Olama & Code GPT
- 10:25NPU models with AI Toolkit
- 13:55GPU LLM performance demo
- 15:42GitHub Copilot review
- 23:01PowerShell scripts with LLMs
- 26:11LLM prompting tips
- 31:39Debugging scripts with LLM assistance
- 34:31Optimizing scripts with LLMs
- 39:50Cloud vs local LLM
- 42:59Q&A
related talks
37mEmpowering Developers with local AI
01h 01mUnlocking the Potential of AI
44mThe Science of Effective Prompting
46mSupercharging PowerShell with LINQ
01h 51mShaping the future of work with AI
46mPowerShell as an MCP Tooling Host: Using MCPServerPS to Supercharge AI-Assisted...
43mBuild and Maintain your Company Copilot with Azure ML and GPT-4 Q&A
27mQ&A #2