Surprised to see Django presented as state of the art... Especially for ML applications, I'd go with Streamlit, Gradio, or Dash. Also, Django has issues when used behind reverse proxies, which is often a requirement for production grade apps.
So, the main reason Django is being presented over others that you mentioned is because:
1.) Django has full components for the front end, and the back end. This means you can build, and productionize complex ML pipelines. With the other 3, those are meant more for building simple interactive dashboards. Think: Complexity vs Simplicity. I've seen several customer facing ML pipelines being made in Django.
2.) Since Django is meant for creating full productionized web apps, it's designed to be scalable from the ground up, and handle huge amounts of traffic. If you are building a customer facing ML pipeline, you definitely want this.
3.) Another reason why I bring Django as state of the art is because it's been around since mid 2000s. Basically, it's had it's "torture test" to see if can be cracked, broken, and it has had several improvements.
You are right, maybe calling it state of the art may not be the best, but rather, calling it the current meta for pulling data. I'll rename the title to be more accurate.
No idea on the reverse proxies aspect, that side is usually handled by the Data Eng, or the Web Dev team, when they are setting this up.
Surprised to see Django presented as state of the art... Especially for ML applications, I'd go with Streamlit, Gradio, or Dash. Also, Django has issues when used behind reverse proxies, which is often a requirement for production grade apps.
So, the main reason Django is being presented over others that you mentioned is because:
1.) Django has full components for the front end, and the back end. This means you can build, and productionize complex ML pipelines. With the other 3, those are meant more for building simple interactive dashboards. Think: Complexity vs Simplicity. I've seen several customer facing ML pipelines being made in Django.
2.) Since Django is meant for creating full productionized web apps, it's designed to be scalable from the ground up, and handle huge amounts of traffic. If you are building a customer facing ML pipeline, you definitely want this.
3.) Another reason why I bring Django as state of the art is because it's been around since mid 2000s. Basically, it's had it's "torture test" to see if can be cracked, broken, and it has had several improvements.
You are right, maybe calling it state of the art may not be the best, but rather, calling it the current meta for pulling data. I'll rename the title to be more accurate.
No idea on the reverse proxies aspect, that side is usually handled by the Data Eng, or the Web Dev team, when they are setting this up.