Google voice intergration into the existing django project

Please build a django app of one page, and help me install ASGI and uWSGI on my ubuntu EC2 server. Requirements 1. On the browser, including mobile ones, users will stream voice to Google Speech-to-text api and receive the recognized text stream in real-time. According to my research, I think the django channel module will do the trick. The users will see the varying, live text from the server, and you can demo how the users will see the varying, live text by doing test here - [login to view URL] 2. At the same time, I also need to record the user's voice so that user can later play back his/her own voice. 3. We have 1 min limit in stream voice recognition, and the live stream will stop after the given second that is less than 60. Also, the stream will stop once the 'practice word' is found on the recognized text stream. 4. My intention is to provide live feedback to user whether specific word is being recognized. This is the language learning app, so I will give a task to user to read, and provide feedback. Once I receive the code, I will attach the business logic in the channel layer. I would appreciate if you can teach me how to receive the text-variable at the channel and process whether the given text is in the recognized text stream. On my app, each page will have different different vocabularies. So some page's task is to recognize 'apple' while another is to recognize 'banana'. Such variable can be best attained at the channel level. Ultimately when the given variable is in the text stream, I want to display "good job" on the webpage and send ajax post signal to log. I know how to send ajax.) 5. Assuming that this job would require django channel, ultimately I need your help to install nginx, ASGI, and uWSGI to make this app run on my existing project on EC2 ubuntu. FYI, currently, it only runs with 'python3 [login to view URL] runserver'. So far I found this article regarding such server structure, but I trust your recommendation if you have any. ([login to view URL])