Summarizing a human conversation has lots of inherent benefits in almost every use-case and domain because of the core value that it brings – shortening the time to grasp the contents of a conversation and saving vast amounts of time.

Symbl.ai, a developer platform for deriving Artificial Intelligence from an analysis of conversation data, is releasing a brand new product called under its experimental wing of “Labs”, for summarization: the Summary API. The API provides developers with the ability to summarize conversation into summaries. The Summary API synthesizes vast amounts of conversation data into a summary.

Automatic summarization is a huge challenge for data scientists. AI models for summarization are required to filter domain-specific language in source texts (scientific research papers, articles, news, legal filings, conversations, or live events). AI models train for the identification of the prominent features in contextually relevant source texts to produce a consistent summary.

Symbl.ai’s Summary API provides a solution to this challenge. The API seeks the most relevant information based upon the context of the conversation. In line with its product suite for generating contextual insights such as follow-ups, questions or action-items, topics and subtopics, or personalized intents, Symbl.ai’s Summary API is an API. It is designed specifically for developers to utilize in their stacks for data science analysis, application development or enhancing real-time communication, engagements or streams.

If you would like to make a call to our Summary API, there are three steps. The first step is to sign up on our platform. The second step is to upload a conversation through our Async APIs with a query parameter for summarization set: `?enableSummary=true`. The third step is to make a call to our Summary API with the conversation ID you obtained from the Async: https://api-labs.symbl.ai/v1/conversations/{{conversation_id}}/summary. If you are unfamiliar with how to make API calls with `cURL` or a programming language like JavaScript or Python, Symbl.ai recommends you check out its collection or public workspace on Postman. Depending upon your set up, you run your APIs calls on a Postman desktop agent or its apps. Symbl.ai’s Postman public workspace is available here.

Currently, the Summary can be enabled with Async APIs. Support for Summarization in real-time will be added soon. However, if you are using Streaming or Telephony API, you can use the refresh parameter (i.e., `?refresh=true`) to generate the Summary after the conversation has ended.

Avatar photo
Eric Giannini
Lead Developer Evangelist & Advocate