Automated content creation and content services for collaboration platforms
Embodiments described herein relate to systems and methods for automatically generating content, generating API requests and/or request bodies, structuring user-generated content, and/or generating structured content in collaboration platforms, such as documentation systems, issue tracking systems, project management platforms, and other platforms. The systems and methods described use a network architecture that includes a prompt generation service and a set of one or more purpose-configured large language model instances (LLMs) and/or other trained classifiers or natural language processors used to provide generative responses for content collaboration platforms.
1 . A computer-implemented method of providing content for a chat session with a user, the method comprising:
receive a request message generated at a first client device by the user, the request message including a natural language user input provided to a chat interface of a messaging platform;
analyze the natural language user input to determine an intent metric;
in response to the intent metric satisfying a criteria, forward the request message to a first recipient;
in response to the intent metric failing to satisfy the criteria, forward the request message to a second recipient operating an automated chat service, wherein the automated chat service is configured to:
generate a first prompt comprising:
first predetermined query prompt text; and
at least a portion of the natural language user input;
communicate the first prompt to a generative output engine;
receive a first response from the generative output engine; and
conduct a search of a knowledge base platform using the first response from the generative output engine,
in response to receiving a set of search results, each result identifying a respective knowledge base content item, generate a second prompt comprising:
second predetermined query prompt text; and
content extracted from at least one knowledge base content item identified in the set of search results;
communicate the second prompt to the generative output engine;
receive a second response from the generative output engine; and
generate a response message comprising at least a portion of the second response from the generative output engine and at least a portion of the set of search results;
in response to the automated chat service generating the response message, provide the response message for transmission to the first client device via the messaging platform.
2 . The computer-implemented method of claim 1 , wherein:
the automated chat service is a first automated chat service;
the first recipient operates a second automated chat service, wherein the second automated chat service is configured to:
analyze the natural language user input to determine a request type; and
select a predefined chat sequence in accordance with the determined request type, the predefined chat sequence defining a series of response entries; and
in response to the selection of the predefined chat sequence, cause a first response entry of the series of response entries to be transmitted to the first client device via the messaging platform.
3 . The computer-implemented method of claim 2 , wherein the intent metric indicates a correlation of the natural language user input with respect to the request type of a set of request types associated with the second automated chat service.
4 . The computer-implemented method of claim 1 , wherein the second response received from the generative output engine includes a content summary generated using first content extracted from a first knowledge base content item identified in the search results and a second content extracted from a second knowledge base content item identified in the search results.
5 . The computer-implemented method of claim 4 , wherein the first content is a block of text within the first knowledge base content item having a semantic similarity with respect to the first response.
6 . The computer-implemented method of claim 4 , wherein the response message comprises:
a first portion including the content summary generated using the content extracted from the at least one knowledge base content item; and
a set of links to at least a subset of the set of search results, each link selectable to cause redirection to a respective knowledge base content item.
7 . The computer-implemented method of claim 1 , wherein:
the first prompt is communicated to the generative output engine using an application programing interface (API) call transmitted to the generative output engine;
the first prompt is a first serialized data object provided to the generative output engine using the API call;
the second prompt is communicated to the generative output engine using the API call; and
the second prompt is provided as a second serialized data object to the generative output engine using the API call.
8 . The computer-implemented method of claim 1 , wherein the response message is transmitted to an operator in communication with the second recipient before being relayed to the first client device via the messaging platform.
9 . A system for providing content for a chat session with a user, the system comprising:
a backend of a collaboration platform configured to provide technical assistance, the backend of the collaboration platform executing on a server having a processing unit and computer readable memory storing instructions for:
in response to receiving a request message generated by a first client device of the user operating a messaging platform, forward the request message to a recipient operating an automated chat service, wherein the automated chat service is configured to:
generate a first prompt comprising:
first predetermined query prompt text; and
at least a portion of a natural language user input extracted from the request message;
communicate the first prompt to a generative output engine;
receive a first response from the generative output engine; and
conduct a search of a knowledge base platform using the first response from the generative output engine,
in response to receiving a set of search results, each result identifying a respective knowledge base content item, generate a second prompt comprising:
second predetermined query prompt text; and
content extracted from at least one knowledge base content item identified in the set of search results;
communicate the second prompt to the generative output engine;
receive a second response from the generative output engine; and
generate a response message comprising at least a portion of the second response from the generative output engine and at least a portion of the set of search results;
in response to the automated chat service generating the response message, provide the response message for transmission to the first client device via the messaging platform.
10 . The system of claim 9 , wherein the instructions further comprise instructions for:
in response to receiving the request message, analyze the natural language user input to determine a content metric;
in response to the content metric failing to satisfy a criteria, forward the request message to the recipient, the recipient being a first recipient; and
in response to the content metric satisfying the criteria, forward the request message to a second recipient different than the first recipient.
11 . The system of claim 10 , wherein:
the automated chat service is a first automated chat service;
the second recipient operates a second automated chat service, wherein the second automated chat service is configured to:
analyze the natural language user input to determine a request type; and
select a predefined chat sequence in accordance with the determined request type, the predefined chat sequence defining a series of response entries; and
in response to the selection of the predefined chat sequence, cause a first response entry of the series of response entries to be transmitted to the first client device via the messaging platform.
12 . The system of claim 9 , wherein the response message comprises:
a first portion including a content summary generated using the content extracted from the at least one knowledge base content item; and
a set of links to at least a subset of the set of search results.
13 . The system of claim 9 , wherein the response message is transmitted to an operator before being relayed to the first client device via the messaging platform.
14 . The system of claim 11 , wherein the content metric is an intent metric indicating a conformity of the natural language user input with respect to a one or more of request types of the second automated chat service.
15 . A computer-implemented method of providing technical assistance using a messaging platform, the method comprising:
receive a user input extracted from a request message generated at a first client device by a first user using the messaging platform;
forward the user input to an automated chat service, wherein the automated chat service is configured to:
generate a first prompt comprising:
first predetermined query prompt text; and
at least a portion of the user input;
communicate the first prompt to a generative output engine;
receive a first response from the generative output engine; and
conduct a search of a content platform using the first response from the generative output engine,
in response to receiving a set of search results, each result identifying a respective content item, generate a second prompt comprising:
second predetermined query prompt text; and
content extracted from at least one content item identified in the set of search results;
communicate the second prompt to the generative output engine;
receive a second response from the generative output engine; and
generate a response message comprising at least a portion of the second response from the generative output engine and at least a portion of the set of search results;
in response to the automated chat service generating the response message, provide the response message for transmission to the first client device via the messaging platform.
16 . The computer-implemented method of claim 15 , wherein the response message includes:
a content summary generated using the content extracted from the at least one content item; and
a set of links to at least a subset of the set of search results.
17 . The computer-implemented method of claim 16 , wherein the content summary is generated using first content extracted from a first knowledge base content item identified in the search results and a second content extracted from a second knowledge base content item identified in the search results.
18 . The computer-implemented method of claim 17 , wherein, the first content is a subset of content of the first knowledge base content item selected based on a semantic similarity with respect to the first response.
19 . The computer-implemented method of claim 15 , wherein the response message is provided to an operator before the response message is transmitted to the first client device via the messaging platform.
20 . The computer-implemented method of claim 15 , wherein, the second response received from the generative output engine includes a content summary generated using first content extracted from a content item identified in the set of search results and second content extracted from a second knowledge base content item identified in the set of search results.