Sometimes everything fits together. The requirements, the tools, the platform. And sometimes it works, but not entirely smoothly. On Power Platform we can build a lot of things. Canvas apps, Power Automate, Dataverse – we know all of these inside out. But just as a developer eventually realizes that one programming language is not enough for everything, the same applies to building in a single ecosystem. There are cases where a Microsoft solution can handle the task, but not without compromises. And that is exactly when it makes sense to look beyond.
Real-World Example: Invoice Processing with Azure AI Document Intelligence
The customer wants to automatically extract data from PDF invoices, process them, and connect everything to their internal systems. Sounds like a perfect case for Azure AI Document Intelligence—at first glance.
It works well, though getting there is not always straightforward.
- you need to train a custom model on sample documents (Custom Model)
- form-based logic sometimes isn’t enough and needs additional fine-tuning
- integration into a canvas app requires a middleware layer using Power Automate or Azure Functions
As with other solutions on the market, results still require a certain level of validation. Microsoft provides a robust platform with a high degree of flexibility and scalability, which is a strong advantage if a company plans to use Document Intelligence across multiple processes. On the other hand, for organizations that prioritize faster deployment, lower complexity, and easier fine-tuning, it may make sense to explore alternative options.
What if there’s a better way?
In cases like this, we look beyond Microsoft. There are specialized OCR services like Rossum, Klippa, or Make that offer ready-made invoice solutions. They support more formats, tend to be more reliable, and often provide direct integration with third-party systems.
And sometimes, something simpler does the trick—a PHP script using Tesseract OCR or a lightweight Node.js service that does one thing, but does it well.
Solution: Not only invoices, but also complex documents
With Azure AI Document Intelligence, we are not limited to extracting data from invoices. The service can also handle more complex documents such as technical specifications or other structured formats. It automatically detects and extracts text, key values, tables, and the overall document layout. This can be achieved through pre-built models, for example for invoices, or through custom models that can be quickly adapted using just a few document samples. The result is precise, reliable, and repeatable extraction even from highly complex documents.
Why It Matters
Because we’re not selling “a Microsoft solution.” We’re delivering a working result—one that makes sense in the customer’s environment and context. If they prefer Microsoft, we’ll build it on the Power Platform. If they’re looking for something more open or cost-effective, we’ll offer a different path.
And it’s not just about the tech. It’s about trust. When a customer feels we’re genuinely recommending the best possible solution—not just whatever’s at hand—it builds respect and partnership. And that’s the foundation of any successful collaboration.
We’re Power Platform developers. But just like a canvas app sometimes needs a touch of CSS, a modern project often needs a broader perspective. Because at the end of the day, it doesn’t matter whether it runs as a Power App or a webhook.
“It’s not about whether we use Microsoft or something else. What matters is that the solution works and delivers real value.”
Ready for a change? Write to us at customers@evenaut.com and let’s take the first step toward a more efficient future for your business.
