PC8 is due Wednesday, 4/29 at 12pm Eastern Time.
The Final Turn-In is a culmination of all the work you have done this semester. You must turn in (1) your front-end source code, (2) your back-end source code, and (3) your platform setup. In addition, you should document these pieces well enough that we can reproduce your conversational AI's setup.
Your Final Turn-in will consist of all the source code, data, and documentation you have worked on this semester. It is meant to be a complete, final version of your project that you might ship to an interested customer.
You must also document your project. For each component (your frontend, backend, and platform), please describe the steps required to setup and launch your project. We want to be able to reproduce your projects (consider demonstrations for future iterations of the course). What resources to we need? Do we need an Amazon EC2 instance? Do we need to use particular network ports for each components? Do we need to upload a file to DialogFlow? Consider the steps required to explain to a tech-savvy engineer how they could remake your project.
Because teams may use different platforms (Clinc vs. DialogFlow vs. Rasa), you are expected to submit your platform as appropriate. For example, Clinc allows you to directly export AI versions as .tar.gz files, while DialogFlow allows exporting as a .zip file. Those can be directly submitted. For Rasa projects, you can submit a .zip file containing the domain spec, the example data, Action Server code, and any other files that make up your Rasa virtual assistant.
When you deliver a product to a customer, it's not enough to have a project that works for you. You must spend time thinking about generalizing to other platforms. What happens when you use a different network address? What happens when someone runs the script in a different directory? What if they have a different set of packages and libraries installed? You must think about the steps required to reproduce the project you've built from scratch on a new system.
Sometimes, people will use virtual machine images or new Amazon EC2 instances to create an example environment in which to attempt rebuilding from scratch. You might consider something similar to help you remember all of the dependencies you have (e.g., how many times did you have to run pip install ... before the thing worked?).
You will turn in a single teamX.zip file consisting of three more zip files:
PC8 Grading (out of 6 points):