CLIENT HAND-OFF PLAN – 3D ENGAGEMENT CUBE

1. Purpose of This Plan

This document explains how we will deliver the finished 3D Engagement Cube system to the client and ensure they can fully operate it on their own machines.

It is not a user manual. It is the plan (required by the course) that shows we have a clear, realistic path to hand off a working system the client controls.

2. What We Are Delivering

We are delivering two main components:

2.1 Machine Learning Scoring Pipeline (Python)

• Runs fully on the client’s local machine

• Reads data from AWS S3 or local CSV files

• Generates Interest, Intent, and ICP scores

• Outputs an easy-to-read CSV file

• Controlled using simple Makefile commands (documented in README)

• Requires only Python and basic technical skill

2.2 3D Visualization Interface (Flutter Web App)

• Allows the client to explore accounts in an interactive 3D cube

• Runs locally on the client’s computer in a Chrome browser

• Can be deployed to any static host (Netlify, Vercel, AWS S3 Website Hosting)

• Does not require a backend server

2.3 Documentation

• README.md (how to run the model, load data, and install dependencies)

• ADMIN_SETUP.md (AWS connection)

• FLUTTER_UI.md (Flutter set up)

2.4 Training Videos

Two required training videos will be delivered:

• Video 1: Installing and setting up the system

• Video 2: Running the scoring pipeline and launching the UI

3. How the Client Will Obtain the System

3.1 GitHub Repository (Primary Delivery Method)

The entire project will be handed off through a GitHub repository.

The client will simply:

1. Open the provided GitHub link

2. Click “Clone”

3. Use the repository directly

The repository will contain:

• All Python code

• All Flutter UI code

• Documentation

• Example configuration files

3.2 Email Communication

The client will receive an email containing:

• Direct link to the GitHub repo

• The two training video links

• Optional PDF copies of README and ADMIN_SETUP

3.3 Final Handoff Meeting

The entire system will be installed together with the client on their machine to confirm independence.

4. Client Setup Plan

This is what the client will do during the handoff meeting.

4.1 ML Pipeline Setup

Client will:

1. Install Python

2. Create a virtual environment

3. Install dependencies

4. Configure AWS credentials (if using S3)

5. Run the scoring pipeline

6. Confirm that the output CSV is generated

All steps are documented in plain language in README.md.

4.2 3D UI Setup

Client will:

1. Install Flutter

2. Navigate to the UI directory

3. Run the UI locally in Chrome

4. Load the generated score CSV

5. Interact with the cube visualization

If the client wants to deploy the UI online later, we will explain hosting options (i.e. Vercel), but deployment is not required during handoff.

5. Technical Summary

The system is:

• A local Python program

• A local Flutter web interface

• Not dependent on UNC infrastructure

• Not deployed to any cloud platform by default

• Completely owned and run by the client

No specialized hardware or servers are required.

6. Costs

Required cost: None.

The system runs locally for free.

Optional long-term costs (only if the client wants extra features):

• Netlify or Vercel for UI hosting: free tier available

• AWS EC2 for automated monthly scoring: small monthly cost depending on instance type

The delivered system incurs no ongoing cost.

7. Discussion With Client

We have already confirmed that:

• Client wants a fully local system

• Client has AWS S3 access

• Client can install Python and Flutter

• A live installation meeting is preferred

• No hosting or cloud deployment is required for initial use

8. Handoff Meeting Plan

Meeting Date: Wednesday, December 3

Agenda (60–90 minutes)

1. System Overview (5–10 minutes)

Brief explanation of the ML pipeline, visualization UI, and expected workflow.

2. Live Installation of ML Pipeline (5-10 minutes)

Client installs everything directly on their computer:

• Clone repo

• Install Python dependencies

• Configure AWS

• Run scoring

• Verify outputs

3. Live Installation of 3D UI (10-15 minutes)

Client installs Flutter and launches the UI.

4. Documentation Review (5 minutes)

Walk through README

Explain where to find help.

5. Final Q&A and Confirmation (10 minutes)

Ensure the client can run everything independently.

10. Completion Criteria

The handoff is complete when:

• Client has cloned the repo

• Client has installed dependencies

• Client has run the scoring pipeline successfully

• Client has run the 3D UI successfully

• Client has received training videos

• Client can reproduce all steps independently

11. After-Handoff Support

• Two-week support window after delivery

• One optional follow-up Zoom session

• Minor fixes to documentation if needed