If you have just a couple of prototype devices sitting on a lab bench, then yes you could make them post data straight into DevicePilot - it's very easy to do (just sign-up and you'll see code samples of how to do it with one line of code) and might be a great solution for instantly visualising your device data, debugging it on the lab bench, sharing it with colleagues, etc.
BUT there is (at least) one very important reason why you need to move beyond this super-simple architecture before your devices leave the lab bench: Security! The above solution requires you to put your DevicePilot access key into every device. If you ship those devices into the real world then someone will discover that "pre-shared key" and get access to read and write from your DevicePilot account - bad!
So the right solution is to get your devices to send their data into a Cloud IoT platform first (*). There are lots of choices, including for example AWS-IoT, Azure IoT and Google IoT. These professional cloud IoT platforms all have mechanisms to ensure that every device gets its own security key, so there is no way for a hacker to compromise all devices by attacking just one.
Once the data is flowing securely into the cloud, it's easy to then get it into other services too - such as DevicePilot. For example, with AWS-IoT you can switch on DevicePilot without writing any code at all, just with a few clicks. Or with any other platform you can do it by adding literally just one line of code (the same code that you put into your devices, but this time the security key is safely held within your cloud application).
(*) of course, it doesn't have to be a cloud IoT platform, you could use the PC under your desk and DevicePilot will still work just fine. But for all sorts of reasons we do strongly recommend a Cloud IoT solution for the heart of your architecture (if you'd like to understand why then please just ask us).