Hi PowerShell Team and community,

I wanted to share something I have been building for the past year -- VBAF,
the only complete AI and reinforcement learning framework written entirely
in PowerShell 5.1.

No Python. No dependencies. Runs on any Windows PC with no setup beyond:

Install-Module VBAF -Scope CurrentUser

What it implements from scratch:

• Neural networks with backpropagation (Rumelhart 1986)
• Q-Learning with Q-table (Watkins 1989)
• Deep Q-Networks with experience replay and target network (Mnih 2015)
• Proximal Policy Optimization -- same family as ChatGPT training (Schulman 2017)
• Asynchronous Advantage Actor-Critic (Mnih 2016)
• Multi-agent market simulation with emergent behaviours
• 14 enterprise automation pillars built on DQN

Every algorithm includes comments explaining the mathematics in plain English
with references to the original research papers. The code is the curriculum.

I noticed your 2026 investments post mentions AI integration with PowerShell
as a priority. VBAF demonstrates that serious AI/RL work is already possible
in pure PowerShell 5.1 today -- and could serve as a reference or educational
resource for the community.

GitHub: https://github.com/JupyterPS/VBAF
PSGallery: https://www.powershellgallery.com/packages/VBAF

Happy to answer any questions about the implementation.

Henning -- Roskilde, Denmark