GuidesAPI ReferenceSDKsIntegrationsWhat's New?
Guides

Async API - Video

Suggest Edits

The Python SDK allow you to asynchronously send video conversation data and generate the following:

  • Speech-to-Text (messages)
  • Action Items
  • Questions
  • Topics
  • Follow-ups
  • Members (a list of members in a conversation).

You can also utilize the parameters function to send additional parameters supported in the Async Video API for file and url upload.

Sample Code

import symbl

conversation_object = symbl.Video.process_file(file_path=r'c:/Users/john/Downloads/video.mp4')
print(conversation_object.get_messages())
print(conversation_object.get_topics())

# You can use the same code to generate other insights such as:
# print(conversation_object.get_follow_ups())
# print(conversation_object.get_action_items())
# print(conversation_object.get_questions())
# print(conversation_object.get_members())

🚧

Always prefix "r" before the file location in the code if you are using Windows system. Example: file_path=r'c:/Users/john/Downloads/business_meeting.mp4

Appending Video API

To append a video call to an already processed video file, use the .append function with the conversation_id as shown below:

conversation_object = symbl.Video.append_file(file_path=r'c:/Users/john/Downloads/video.mp4', conversation_id='')

A complete sample of the .append function is given below:

import symbl

conversation_object = symbl.Video.append_file(file_path=r'c:/Users/john/Downloads/video.mp4', conversation_id='')
print(conversation_object.get_messages())
print(conversation_object.get_topics())

# You can use the same code to generate other insights such as:
# print(conversation_object.get_follow_ups())
# print(conversation_object.get_action_items())
# print(conversation_object.get_questions())
# print(conversation_object.get_members())

Using Parameters

In the sample below, the optional parameters such as name, detectPhrases, etc. have been used. You must pass the query parameters as parameters as shown below.

You can use these parameters in the .append functionality as well.

For the complete list of supported parameters, see Process Video. 

import symbl

conversation_object = symbl.Video.process_file(file_path='/users/jon/Downloads/Welcome.mp4', 
parameters={
    'name':'Business Meeting', 
    'detectPhrases': True, 
    'enableSpeakerDiarization': True, 
    'diarizationSpeakerCount': 3, 
    'channelMetadata': [
      {"channel": 1,"speaker": {"name": "Jon Snow","email": "[email protected]"}}
     ]})
print(conversation_object.get_messages())
print(conversation_object.get_topics())

# You can use the same code to generate other insights such as:
# print(conversation_object.get_follow_ups())
# print(conversation_object.get_action_items())
# print(conversation_object.get_questions())
# print(conversation_object.get_members())

Async Video URL API

Use the code given below to process video calls with the Python SDK and generate conversation insights such as speech-to-text, topics, follow-ups, action items, questions. It can be utilized for any use case where you have access to recorded video stored publicly as a URL and want to extract insights and other conversational attributes supported by Symbl.

📘

The URL provided must be a publicly available URL. Currently we do not support any redirected links, shortened links (e.g. bit.ly), YouTube, Vimeo, or links from any audio/video platforms.

import symbl

request_body = {
   'url': ‘https://symbltestdata.s3.us-east-2.amazonaws.com/sample_video_file.mp4’,
   'name': 'Business Meeting',
 }

conversation_object = symbl.Video.process_url(payload = request_body )
print(conversation_object.get_messages())
print(conversation_object.get_topics())

# You can use the same code to generate other insights such as:
# print(conversation_object.get_follow_ups())
# print(conversation_object.get_action_items())
# print(conversation_object.get_questions())
# print(conversation_object.get_members())

Appending Video URL API

To append an already processed video file via URL, use the .append function with conversation_id as shown below:

conversation_object = symbl.Video.append_url(payload=request_body, conversation_id='4639962491256832')

A complete sample for the .append function is given below:

import symbl

request_body = {
   'url': ‘https://symbltestdata.s3.us-east-2.amazonaws.com/sample_video_file.mp4’,
   'name': 'Business Meeting',
 }

conversation_object = symbl.Video.append_url(payload=request_body, conversation_id='4639962491256832')
print(conversation_object.get_messages())
print(conversation_object.get_topics())

# You can use the same code to generate other insights such as:
# print(conversation_object.get_follow_ups())
# print(conversation_object.get_action_items())
# print(conversation_object.get_questions())
# print(conversation_object.get_members())

Using Parameters

In the sample below, the optional parameters such as name, detectPhrases, etc. have been used. The supported parameters must be sent as a part of the request_body function.

You can also use parameters while using the .append functionality.

See the complete list of supported parameters here. 

request_body = {
     'url':'https://symbltestdata.s3.us-east-2.amazonaws.com/sample_video_file.mp4', 
     'detectPhrases': True, 
     'channelMetadata': [{"channel": 1,"speaker": {"name": "Jon Snow","email": "[email protected]"}}]
     }
 conversation_object = symbl.Video.process_url(payload=request_body)
 print(conversation_object.get_topics())
 print(conversation_object.get_messages())
# You can use the same code to generate other insights such as:
# print(conversation_object.get_follow_ups())
# print(conversation_object.get_action_items())
# print(conversation_object.get_questions())
# print(conversation_object.get_members())

Utilizing the wait Parameter

Use the wait parameter (by default set to True) while making concurrent API calls. Setting the value wait=False will execute an existing function on a separate thread making it a non-blocking API call. It also has the callback support.

Example:

conversation_object = symbl.Video.process_file(file_path=r'c:/Users/john/Downloads/video.mp4', wait=False)

Python SDK Reference

For a complete list of supported classes and objects in the Python SDK, see the Python SDK Reference page.

You can view more capabilities added to Async API in the following sections:

Updated almost 2 years ago