Proj 2: Location Strategy

Business Problem: 

You work for a Chicago financial services company focused on investments and retirement planning for individual investors.  Your company is expanding and wants to open new offices in desirable cities.   Your task is to extract, merge, transform, and analyze five location-related datasets to come up with a scoring formula to rank-order the top 500 cities in the country. 

Requirements: 

This is a group project. You will need to document your process, capturing what you do, why you did it, and who did it, along with your decision-making process. You will:

1. Put all your documents in a Google Drive folder for your group.  Make sure that the folder is shared to be world-readable. 

2. Capture your work in two Custom GPTs / AI agents that allow interactive exploration and explanation of your work for other people who also have a paid ChatGPT Plus subscription.  

3. Create a detailed web page summary on your website that can be used in a job interview to describe this project. Include links to documents and the AI agents.

Details: 

This project will be graded out of 52 points and then scaled for the number of points it counts towards your final grade. 

Documents (30 points): 

You will need to create and include the following ten documents. All these documents must be in your shared Google Drive folder for your group.  These documents should describe:

1. The division of labor in your group, describing who did what, and how you made those decisions.

2. The process you went through, capturing the order of what you did, problems encountered along the way, and how you used AI along the way

3.  For each of the six datasets (so six documents in total):

   •  The Extract/Transform/Load (ETL) work to combine and select features, explaining why some features were selected for use and others were excluded, and:

   • The Exploratory Data Analysis (EDA) for all selected features, including summary statistics, relevant correlations, and descriptions of actions taken to handle missing values and transformations, and any relevant visualizations.

4. Your scoring formula for rank-ordering cities, including features and weights per feature.  Your formula must include a minimum of seven features. How did you come up with this formula?  How does it compare to other such "Best Cities" work that has been previously done elsewhere? How did you use AI to get this done?

5. A highlights article as a Google doc (or pdf, added 10/15), showing the five best and the three worst cities according to your analysis, including pictures, descriptions, and intuition on the rankings. This highlights document should be presented like a magazine article, with appropriate visual components.

Evaluation Rubric: (3 points for each of the ten documents, so 30 pts total).  Each document will be graded as:
        0 points: Not done
        1 point:   Significantly incomplete
        2 points:  Almost everything is there
        3 points:  Complete and full descriptions. Includes appropriate visual elements.

Custom GPTs / AI Agents (12 points): 

Create two custom ChatGPTs / AI Agents:

1. A custom GPT / AI agent to help you choose the best features from each dataset and weights for the selected features. This agent should include the assignment description, a data dictionary for each dataset, and the role you want it to play.  This agent should be shared, with a link provided near the top of your Project 2 web page.

2.  A custom GPT / AI agent to summarize your project.  This custom GPT / agent should include all your analysis and final documents and a description of the role you want it to play.  It should allow the user to ask questions about the respective roles that different group members took, your group's process, your work with the datasets, how you chose the features and weights for your rank ordering, and extra contextual information about the best and worst cities you selected.   Again, this agent should be shared, with a link provided near the top of your project 2 web page.
   (Announced in class 10/6 and added here by Dale 10/7:)  This Custom GPT / Agent will be the basis for the live demo you give as your presentation of this project.

Evaluation Rubric: (3 points each, 12 points total)  (Rubric below updated 10/7 by Dale)   Each agent will be graded on a scale between 0 (not done) and  7 (agent is complete and answers all questions well).     Each scoring item below will be given 0..3 points each (12 points total) as follows: 

        0 points: Not done
        1 points: Partially done
        2 points: Mostly done, with some issues present
        3 points: Done completely, with all aspects addressed

Scoring Items;

1. Link to agent provided near the top of the project 2 web page

2. Custom GPT / Agent description is appropriate, indicating the persona it should take on and the role it should play

3. Appropriate files and other resources are included with the agent

4. Agent works as intended,  taking on the described role and reflecting the context given by the included files and resources
   
   

Web Page (10 points): 

This page should serve as an ideal landing point for a resume link describing this project as a STAR story (Situation, Task, Action, Result) reflecting your personal contributions to the project. At the top of your web page (as previously mentioned), you must have links to the publicly-readable Google Drive folder as well as to both of your AI agents. 

Evaluation Rubric: (2 points each, 10 points total) Each scoring item below will be given 0..2 points each as follows: 

        0 points: Not done
        1 points: Partially done
        2 points: Substantially or completely present
Scoring Items:

1. 1.  Is explicitly structured as a STAR story, detailing your personal contributions and learning 

   2. Sets the stage and explains the background situation for the problem

   3. Describes the task you had to do

   4. Details the actions taken by the group, highlighting your individual contributions

   5. Presents the results of what you accomplished

Turning it In: 

You will "turn in" your work by posting it on your website on the Project 2 page you created previously. We will grade your group's work by selecting a random group member and using their web page entry, Google Sites link, and AI agent to represent your entire group.  Be sure to check each other's websites, since you don't know whose will be chosen!