PC7 — Final Presentation and Demonstration
PC7 is due Monday, 4/20 at noon.
You will work with your team to deliver a live summary presentation of your project.
You will also deliver a live demonstration of your project.
Finally, you will deliver a poster that summarizes your project.
Please schedule a 30-minute block with your group members using this Google Calendar link.
Specification
The Final Presentation will be conducted as a remote meeting with the instructional staff. Your task is to present the culmination of your project to an audience that is unfamiliar with Conversational AI.
You will deliver a presentation with your group that assumes the audience knows nothing of Conversational AI techniques. Your job is to convince such an audience that your project solves a problem, and explain how your Conversational AI project works at a high level.
Your final presentation should tell us a problem, its scope, and
then introduce concepts from Conversational AI that you have learned
this semester (how do we model conversations?). The presentation
should tell us why a virtual assistant can help solve your problem.
You can describe steps like crowdsourcing, amount of training data,
etc. The presentation itself should not exceed 10
minutes.
In addition, you will give a live demo of your project (separate
from the 10-minute presentation). Depending on your group, you could
make this a separate video that we see during the remote meeting.
To the best of your abilities, please plan to make it a live
demo conducted as part of the remote meeting.
As with the other presentations this semester, not every student in
your group needs to actually deliver the presentation. However,
please try to attend your group's final presentation.
There will also be questions at the end of your presentation.
Finally, you will turn in a PDF of a poster that summarizes your
project. Your poster should contain (1) a description of the problem,
(2) a state graph diagram, (3) a listing of intents/slots and the
volume of data, and (4) example interactions with your system. You can definitely include extra material where appropriate, such as screenshots or findings from testing.
Resources
Example Poster 1
Example Poster 2
There is no suggested slide deck for your final presentation. We
want you to decide for your group what elements you think are
important for clearly communicating a relevant problem, your
understanding of conversational AI, how a virtual assistant solves a
problem, and a compelling demonstration that your project can solve
that problem.
You could discuss a Problem Statement, cite relevant
news articles, show screenshots or photos of how the problem
manifests, etc. to motivate your project. You could also discuss state graphs, intent
classification, etc. to demonstrate your understanding of
Conversational AI.
Commentary
This part of the project is meant to show off the highlights of your project that you've worked on this semester. However, we're also using this assignment as a means of gauging your understanding of Conversational AI. Consider: if you're trying to pitch a project involving your virtual assistant, you should be comfortable demonstrating your expertise with Conversational AI as a topic. We're looking for that understanding in addition to evaluating the quality of the project overall.
Hints
I will be taking on the role of someone who knows nothing of Conversational AI during your presentation. I am trying to simulate a stakeholder that is distant from the technology but that is interested in the space. I am looking for your group to convince me that you understand Conversational AI topics, AND that your group's project solves a tangible, relevant problem.
What To Turn In For PC7
You will turn in a single .zip file containing only two files:
- teamX.pdf, containing your group's pitch slides.
- posterX.pdf, containing your group's poster slides.
Grading Rubric
PC7 Grading (out of 30 points):
- 5 points — Presentation oration
- 5 points for clear, projected speech that is demonstrably rehearsed and presented in a professional manner
- 3 points for good speech, not too much stumbling
- 1 point for unclear speech, unrehearsed, questionable organization, lack of professionalism
- 0 points for poor, low speech, no engagement with audience, poor time management
- 5 points — Project introduction
- 5 points for introducing the problem and demonstrating the need to solve it. It is clear why the solution you are proposing solves the issue.
- 3 points for introducing the problem and demonstrating a need to solve it. The setup and the solution is adequate but it is not immediately clear why this is a perfect fit
- 1 point for somewhat introducing the problem. There are major details missing on why this problem needs to be solved and how the product solves it.
- 0 points for not introducing the problem.
- 5 points — Product description
- 5 points for detailing the product and its architecture. The frontend, backend, and conversational AI agent are all mentioned. The discussion details the technical aspects of all 3 components, including the resources used and the conversational AI aspects used. The usage of the product and input/outputs are discussed
- 3 points for detailing aspects of the product. All components are discussed but the technical discussion may not fully explain the architecture of the program
- 1 points for attempting to discuss the product. Not all of the components are discussed and the technical discussion may be lacking.
- 0 points for no discussion of the product
- 5 points — Demo
- 5 points for demos that demonstrate a fully working product. All features are demonstrated and it is clear how to use the product. The product clearly solves the problem demonstrated. Minor bugs may be present but they don’t take away from the demonstration
- 3 points for demos that demonstrate a mostly working product. Most features are demonstrated and it is clear how to use the product. The product clearly solves the problem demonstrated. Some bugs may be present
- 1 point for demos that demonstrate a product that may not work. A few features are demonstrated, but it isn’t clear how to use the product. There are major bugs or lack of functionality
- 0 points for no demo
- 10 points — Poster — You will submit a poster detailing your product. This takes the place of the poster session that would have otherwise occurred in the BBB atrium at the end of the semester. This should again be oriented so that people without a technical background can understand what work has been done while people with a technical background can also benefit from it.
- 10 points for a poster that contains a description of the project, a state diagram, a listing of features such as intents and slots, example utterances and interactions, and details about the conversational AI agent implemented.There is enough detail so that a person with no technical experience can understand. The architecture of the program is discussed, including the front end and the backend.
- 8 points for a poster that contains at least 4 of the 6 components below: a description of the project, a state diagram, a listing of features such as intents and slots, example utterances and interactions, details about the conversational AI agent, the architecture of the program including the front end and the backend. Most of the details are at a level for someone with no background in conversational AI can understand.
- 6 points for a poster that contains at least 3 of the 6 components below: a description of the project, a state diagram, a listing of features such as intents and slots, example utterances and interactions, details about the conversational AI agent, the architecture of the program including the front end and the backend. Most of the details are at a level for someone with no background in conversational AI can understand.
- 4 points for a poster that contains at least 2 of the 6 components below: a description of the project, a state diagram, a listing of features such as intents and slots, example utterances and interactions, details about the conversational AI agent, the architecture of the program including the front end and the backend. Most of the details are at a level that someone with no background in conversational AI will struggle to understand the material.
- 2 points for a poster that contains at least some component below: a description of the project, a state diagram, a listing of features such as intents and slots, example utterances and interactions, details about the conversational AI agent, the architecture of the program including the front end and the backend. Most of the details are at a level that someone with no background in conversational AI will struggle to understand the material.
- 0 points for no poster