Transcribed

Python Is Dead: The AI That Killed Python for Microsoft Automation

Nov 22, 2025 · 26m 4s
Python Is Dead: The AI That Killed Python for Microsoft Automation
Chapters

01 · The Python Dilemma in Microsoft's AI Stack

0s

02 · The Hidden Costs of Python in Power Automate

32s

03 · The Pitfalls of Using Python as Glue

1m 41s

04 · The Power of AI-Assisted Orchestration

3m 52s

05 · Contained Analytics: The Right Place for Python

4m 29s

06 · The Manual Coding Loop: A Recipe for Disaster

4m 48s

07 · The Agent-Driven Approach to Orchestration

7m 10s

08 · Power BI Data Flows: Python's Proper Place

12m 42s

09 · Power Automate: Replacing Python with Office Scripts

15m 51s

10 · Fabric Notebooks: Containing Python in Analytics

19m 11s

Description

Many teams push Python into Power Automate and Power BI as glue and then fight fragile flows, drifting dependencies and scattered logs. This episode shows Power Platform makers and architects...

show more
Many teams push Python into Power Automate and Power BI as glue and then fight fragile flows, drifting dependencies and scattered logs. This episode shows Power Platform makers and architects how Copilot, Office Scripts and agents with typed contracts take over the orchestration while Python stays where it shines: analytics.

In this episode
  • 00:00 – AI writes the glue, not you
  • 00:33 – Python's friction inside Power Platform
  • 05:01 – Why manual code and static playbooks fail
  • 07:36 – Agent plus Copilot as the orchestrator
  • 12:57 – Power BI dataflows: generated, not hand-written
  • 16:06 – Power Automate: Office Scripts plus agents
  • 19:23 – Fabric notebooks: contain Python
  • 23:37 – 30-day implementation playbook

Key takeaways
  • Inside Power Platform, Python is indirect glue that adds external compute, secrets, cold starts and version drift, while Office Scripts and native connectors run where the Microsoft 365 data lives.
  • Let Copilot and agents generate scripts and Power Query steps against typed contracts and the semantic model, then review the diff and keep the prompt as a reusable asset.
  • Keep Python contained in Fabric notebooks and dataflows for analytics kernels, and let the platform and agents handle scheduling, validation, I/O and approvals.

Worth quoting

“office scripts aren't toy typescript. They are the sanctioned edge for automating Excel and related Microsoft 365 tasks without dragging in infra.” – Mirko Peters (02:38)

“Dynamic typing's freedom is fun until it's 3 p.m. on quarter close and your flow succeeds with the wrong shape. Strong typed boundaries catch stupidity before it ships.” – Mirko Peters (03:48)

Explore more on M365.FM
  • Episode page: the full episode with chapters, quotes and a 30-day migration plan
  • Power Automate – where Python glue behind custom connectors makes flows brittle; Office Scripts and native connectors replace it
  • AI agent – agents with typed tool contracts orchestrate Office Scripts, dataflows and notebooks instead of hand-written glue
  • Fabric notebooks – where Python stays contained as the analytics kernel, called by agents through typed job inputs and outputs
  • Related episode: Power Platform Copilot Integration - Simply Explained

Follow and rate M365.FM in your podcast app, and join the free M365 community at m365con.net.

Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-microsoft-365-azure-cloud-power-platform-ai--6704921/support.
show less
Information
Author Mirko Peters M365, Azure & AI
Organization m365 FM
Website www.m365.fm
Tags

Looks like you don't have any active episode

Browse Spreaker Catalogue to discover great new content

Current

Podcast Cover

Looks like you don't have any episodes in your queue

Browse Spreaker Catalogue to discover great new content

Next Up

Episode Cover Episode Cover

It's so quiet here...

Time to discover new episodes!

Discover
Your Library
Search