Go to https://facebook.com.ph and use these credentials to access the chatbot.
On the homepage shortcuts, click PUrple chatbot. On the page, hover on Send Message button and click 'Test Button'
The idea behind this framework is how we can simplify the integration, creation, and interaction of Facebook chatbot and RPA. This applies to people or companies who rely heavily on-premise solutions, servers, since the setup is mostly on the machine. Although, some components can still be applied on the cloud.
What it does
The framework is a set of standards like RE-Framework, the difference is that the built in processes are made for the chatbot. It covers a lot of major areas such as Session Handling, Response Handling, Usage of RE-Framework and Orchestrator. The end goal of this framework was to interact with the user, having to change sessions upon input requests. The chatbot uses an custom made API which sends a transaction Item to the Queue Service and start a Job Process.
How we built it
We built this using our own resources. For the creation of the framework, we used UIPath Orchestrator, Robot, and Studio that are on Community License mode. In order to connect the framework with the chatbot, we used our own Amazon Web Services (AWS) cloud platform to handle the API requests- since we don't have a background setting up a on-premise api. This API is also connected to Facebook chatbot and uses Orchestrator API.
Challenges we ran into
The challenges we ran to are the following:
- Time constraint: since we are currently working on a company and that we only do this after working hours.
- Resources: most of the resources are self-availed, meaning we had to shell out expense to avail the services (but worth it). Another resource is on the end of developers, since we are composed only of 2 developers.
Accomplishments that we're proud of
We are proud of what we have made, the framework. Even though it is still incomplete, initial, and has a lot of areas to improve on, we are able to pull of one working framework that could handle simple requests.
What we learned
What we learned from this hackathon is how we value time and the resources. We got to discover more of our potential on developing this framework and that most of the stuff that we did here is very valuable and reusable on our future projects.
What's next for RPA Facebook Chatbot Framework
The next ideas for the RPA Facebook Chatbot Framework are to integrate machine learning, handle the other responses, optimize database handling, and improving on the architecture of the framework.