Sambit Paul

So, to limit access to URLs based on users, you will need to use roles for the users. So, there will be groups under which all users need to categorised. Let's say in a Hospital setting, the roles would be "Patients", "Doctors", "Nurses" etc. And all the users will be under these categories. Django has an open source extension "django-role-permissions" which I have used extensively for many of my projects and I can vouch for. It is easy to handle and has good documentation to get started with.

For limiting API calls, you will need to create a table which accounts for the number of times a specific API is called by a user and use that to limit access.

Link to mentioned library: https://github.com/vintasoftware/django-role-permissions

Since what you are planning is small scale, I would recommend using Flask. Django is a "full-stack" web framework which means, although it is highly nuanced with lots of features, it has a very steep learning curve. Flask on the other hand is minimalist and easy to get started with. On the front end aspect, React has always been my preference (fundamentally because it had more community support when I started) but that is because most of my projects use that and I have also been working on React Native so its transferable. But if you are starting afresh, you may also want to take a look at Vue JS which has garnered a lot of popularity from developers and now has a lot of community support as well. This article might help you decide on that.

Amazon SQS is simply a queuing service. It is a producer-driven system in which, the SQS sends the message to a consumer and that is it. The messaging system cannot be driven by the consumer. Amazon MQ (which runs on ActiveMQ), is a message broker which means that, the messaging system is consumer driven. Also, MQ has the scope of variable retention period which can be upto a few days. You can keep track of the message that is received and call back on that from the broker at a later time (as long as that is within the retention period). Once all of this is coded into your consumer, then Amazon MQ can be used for pulling back messages from the broker on demand.