Operationalise your Chatbots with Zero-Coding using Flogo
By:
Last updated:
12:47am Nov 29, 2018

 

Building Chatbots can be an effective way for developers to get acquainted with the world of AI. According to Gartner, 25% of customer service and support operations will integrate virtual customer assistant (VCA) or chatbot technology across engagement channels by 2020. These trends are indicative of the fact that Chatbots will soon become the face of customer service in the future and developing chatbots will continue to increase its potential as a highly sought after skill. However, it's important to remember that Chatbots are not meant to be limited to responsive action alone but should be built around proactively tailoring the customer's experience to improve efficiency. Their responses need to be restructured from being simply from point-to-point to being predictive, contextual and conversational in nature.

 

Operationalise your Chatbots

TIBCO enables Developers to operationalise chatbots using a zero coding approach. Using TIBCO's capabilities, Developers will be able to build autonomous, intelligent bots that have the ability to interact with an organization’s core systems and provide key services to customers for better operational efficiency. These empowered bots can expose a business's core systems to customers without affecting data security and equip them with the ability to drive self-service actions such as user profile updation and order tracking. In addition, the bot's functioning can be further enhanced with the capability to seamlessly leverage the embedded historical data of the customers to provide them with personalized interactions, offers and other contextual services.

 

Architecture

TIBCO Flogo Enterprise deployed on AWS LAMBDA enables Developers to create low code lightweight applications that are based off human interactions and can be seamlessly integrated with any NLP. The NLP identifies a Customer Intent and operationalizes it using our platform. You will be then need deployment flexibility for these applications which facilitates back end functionalities and embeds model driven responses that are routed back to the NLP engines. TIBCO natively supports a Cloud-Neutral approach where our applications are certified to run on any of the major cloud providers like AWS, Azure, Google Cloud etc.

Part 1

Creating your Backend Microservices

You can leverage TIBCO Cloud Integration Platform to create integration applications which can seamlessly facilitate your back end functionalities like authentication, balance verification, interacting with your core systems like SAP, AS 400 and Mainframe systems.

TIBCO makes it very easy for you to develop integration systems that are microservices compliant.

Here’s a screenshot of an application that we have built off of leveraging TIBCO Cloud Integration that authenticates the user and exposes the functionality as a rest API.


 

We have also attached a TCI project called verification_v1 to showcase the projects which has been deployed onto TIBCO Cloud.

Please refer to the documentation at : https://community.tibco.com/wiki/tibco-cloud-integration-getting-started...

 

Once you have the cloud studio installed, please see the attached document giving detailed instructions to get you started with building your microservices using TIBCO.

 

Part 2

Building a Consumable Data Model for your Chatbot
 

You can create your data model using TIBCO and embed it as a rest API so that downstream consumption via the chatbot becomes easy.

The data model can provide real time pricing to customers based on inputs given about age of the User, insurance claim details, date of purchase of plan etc.

Go through the following document to learn more about instructions to attach TIBCO Spotfire SS and deploy it to TSSS.

https://docs.tibco.com/pub/enterprise-runtime-for-R/4.0.1/doc/html/TIB_t...

 

Part 3

Building Light-Weight LAMBDA Functions with Zero Coding

 

TIBCO Flogo Enterprise enables Developers to build microservices with a browser based flow designer making app building a faster and more efficient process. The flows can then be deployed to any infrastructure: on-premises, at the edge in devices, and in web-scale serverless environments. The framework allows for developers to build natively for multiple types of cloud platforms and devices so that a broadest range of deployment scenarios are supported without infrastructure lock-in. With Flogo Enterprise, application sizes are significantly smaller than when built using other technologies, and can start faster and scale more efficiently using FaaS than alternative options such as Java or Node.js.

 

Here is a sample of a flogo application which has been used to build out the Lambda function deployed to AWS. This serves as the Fulfillment for any intent in the Lex framework.


 

To get started with Flogo, please have a look at the documentation below:

https://tibcosoftware.github.io/flogo/getting-started/getting-started-we...

https://tibcosoftware.github.io/flogo/faas/how-to/

 

We have also attached a Policymodel.json file which is a sample of how we invoke the data model deployed in Spotfire. The file is attached in word format but will need to be converted to json before you upload it onto Flogo.

Similarly, the other Flogo based lambda functions have been set up to perform the verification and other operations needed for the Bot function.

 

Conclusion:

TIBCO Flogo Enterprise lets you take advantage of new IoT, edge, and serverless application architectures to collect and analyze data locally on the smallest of small IoT devices so that decisions and actions can be made in real time.

Use the below link to get a personal insight into how TIBCO’s Flogo Enterprise can help you operationalize your chatbots.

https://lex-web-ui-codebuilddeploy-1qx3j39gn-webappbucket-13r0wcaz5nx1a....

Attachments

AttachmentSize
Package icon verification_v1.zip40.12 KB
File policymodel_json.docx13.11 KB