Back to results
US 12,596,991 B2Granted
No-code generation of integration applications
AssigneeSalesforce, Inc.
InventorsIgnacio Manzano, Diego Larralde, Tomas Fernandez Martinez
Life of this patent
Estimating the term…
Abstract
A system, method, and device embodiments are provided for programmatically generating and deploying an integration application based on a natural language request without requiring any coding by a user. The application generator infers the sources, targets, connectors, operations, entities, and data mappings needed to build the requested integration application. An exposed web service or API may receive natural language input, determine the meaning of the request, and generates and deploys the resulting integration application without requiring any coding by a user.
Claims20
A method, comprising:
- exposing a web service that receives a natural language input from a chat application;
- tokenizing the natural language input to determine a source token, an entity token, and an operation token;
- accessing an application programming interface (“API”) catalog to map the source token to an API having an API specification defined in the API catalog;
- determining an entity from the natural language input by accessing the API specification in the API catalog to map the entity token to the entity by comparing the entity token to one or more entities detailed in the API specification to determine a matching entity among the one or more entities using a phonetic algorithm, the entity accessible via the API;
- in response to determining that the operation token does not match one or more operations detailed in the API specification, receiving additional information about the operation token in the chat application and determining an operation from the natural language input using the API specification and the additional information about the operation token;
- generating, by one or more processing devices, an integration application that performs the operation on the entity at a source and sends a result to a target;
- and
- deploying the integration application to an integration platform.
5 more elements
A system, comprising:
- a memory;
- at least one processor coupled to the memory and configured to:
- expose a web service that receives a natural language input from a chat application;
- tokenize the natural language input to determine a source token, an entity token, and an operation token;
- access an application programming interface (“API”) catalog to map the source token to an API having an API specification defined in the API catalog;
- determine an entity from the natural language input by accessing the API specification in the API catalog to map the entity token to the entity by comparing the entity token to one or more entities detailed in the API specification to determine a matching entity among the one or more entities using a phonetic algorithm, the entity accessible via the API;
- in response to determining that the operation token does not match one or more operations detailed in the API specification, receive additional information about the operation token in the chat application and determine an operation from the natural language input using the API specification and the additional information about the operation token;
- generate an integration application that performs the operation on the entity at a source and sends a result to a target;
- and
- deploy the integration application to an integration platform.
7 more elements
A non-transitory computer-readable device having instructions stored thereon that, when executed by at least one computing device, causes the at least one computing device to perform operations comprising:
- exposing a web service that receives a natural language input from a chat application;
- tokenizing the natural language input to determine a source token, an entity token, and an operation token;
- accessing an application programming interface (“API”) catalog to map the source token to an API having an API specification defined in the API catalog;
- determining an entity from the natural language input by accessing the API specification in the API catalog to map the entity token to the entity by comparing the entity token to one or more entities detailed in the API specification to determine a matching entity among the one or more entities using a phonetic algorithm, the entity accessible via the API;
- in response to determining that the operation token does not match one or more operations detailed in the API specification, receiving additional information about the operation token in the chat application and determining an operation from the natural language input using the API specification and the additional information about the operation token;
- generating an integration application that performs the operation on the entity at a source and sends a result to a target;
- and
- deploying the integration application to an integration platform.
5 more elements
Dependencies are inferred from the claim text.