As happens in all development, I flubbed and made a mistake.
I started off building this pretty insane Power Automate Flow that was going to change the world, and near the end, as I was finishing the last few bugs, I realized that perhaps Power Automate wasn’t the tool for the job.
But that happens, you start down a path and want to see it to the end – you want to see if the tool itself, the one you are using, could get the job done.
The problem: Purge data from our development environments that we don’t need, don’t delete all of it, leave a sample set that we can still test scenarios with.
If you’re familiar with QA testing, then you know if you have data over 10 years at 12 months a year, that is 120 points of data you want to validate for historical context. From there, you want about 100 records a month, so that means you are saving about 12,000 records of your total 1 million+ records to save. And the fun part here is that you don’t want your data set to be the last current month; you want it to be realistic, as mentioned above.
So the result was this pretty insane flow where I’m only showing one of 4 screens, which, although clean and tight, was too much for Power Automate to do.
Which invariably resulted in this email, where I should have switched lanes, but was pot committed at that point.
Because even though things go a little slow, she still ran like a dream.
But I didn’t have the right tool for the job.
Enter Azure Logic Apps
What I should have done is build the flow in Azure Logic Apps. The best way to think about Azure Logic Apps is that they are Power Automate’s bigger, cooler, older brother that wants you to code, wants you to do super hard problems, and wants to run whatever you throw at it.
I have yet to migrate my flow over to Azure Logic Apps as from this experiment, I have started to jump more and more into that platform. At first glance, from the designer to the connectors, both platforms look relatively the same, but there are some key differences that Claude will help illustrate.
| Feature | Power Automate | Azure Logic Apps |
|---|---|---|
| Target Audience | Business users, citizen developers, power users | Professional developers, IT professionals |
| Primary Use Case | Business process automation, workflow automation | Enterprise integration, system-to-system integration |
| Interface | Web-based designer, mobile app | Azure Portal, Visual Studio Code, Visual Studio |
| Pricing Model | Per-user or per-flow licensing | Pay-per-execution (consumption) or fixed pricing (Standard) |
| Deployment | Cloud-only (SaaS) | Cloud (Azure), on-premises (via ISE), or hybrid |
| Code Experience | Low-code/no-code focus | Supports both low-code and code-first approaches |
| Version Control | Limited native support | Full support via Azure DevOps, GitHub |
| Connectors | 1000+ connectors (includes premium) | 1000+ connectors (mostly the same as Power Automate) |
| Custom Connectors | Supported | Supported |
| Integration | Deep integration with Microsoft 365, Dynamics 365, Power Platform | Deep integration with Azure services, APIs, B2B scenarios |
| Governance | Microsoft 365 admin center, DLP policies | Azure management and governance tools |
| Execution Environment | Multi-tenant cloud | Multi-tenant, single-tenant (Standard), or ISE |
| Triggers | Scheduled, automated, instant, business process flows | Request, recurrence, event-based |
| AI Builder Integration | Native integration | Available through connectors |
| Development Environment | Browser-based only | Azure Portal, VS Code, ARM templates |
| Monitoring | Built-in run history and analytics | Azure Monitor, Application Insights, Log Analytics |
| Best For | Departmental automation, personal productivity, business workflows | Enterprise-scale integrations, complex B2B scenarios, DevOps workflows |
There are a few things with which I don’t 100% agree that I’ll call out here;
- Power Automate has its own native GIT/Pipeline integration that is slick.
- Let’s not forget Power Automate’s 28-day history that cannot be changed -> yeah for Azure Logic Apps and Insights
- I know many companies that are running enterprise-grade flows beyond personal productivity. Maybe for complex B2B integrations, but at that point, you should be leveraging APIM and Custom Connectors to abstract that logic from your flows.
As time persists, it will not surprise me if Azure Logic Apps moves forward with a deeper Azure AI Services integration vs Power Automate’s AI Builder. Not having to deal with Power Automate’s licensing was a big win for me.
Now let’s go fix a flow.