Back to Top

Labely

I created an NLP-based email analysis tool for company complaints for my final project in 15-112, Fundamentals of Programming and Computer Science at Carnegie Mellon University.

Labely mockup banner

Roles

Full-stack development

Timeline

November 2018

Team

Bennett Huffman

Tools / Methods

Python (using NLTK, Matplotlib, and NetworkX)

Vision

Create an email analysis tool for company complaints.

Check out the project on GitHub

Discover

Overview

I read an article before this project illustrating the thousands of emails sent to political campaigns each year, and how many use template responses to help respond to as many as possible. What if I could create a tool that helps label the emails by topic and help summarize these emails with valuable insight?

That’s why I chose to create Labely, an NLP-based email analysis tool that generates labels for topics and conducts both semantic and sentiment analysis with details and summary insights to help identify important emails.

Define

Basics

Thinking about common problems companies may deal with when sent thousands of emails per day, I defined six features I could build into the program:

Musillow was created for these reasons—to help people fall asleep faster, without the hazards and pitfalls of existing solutions.

Design

Wireframes

I sketched my ideas onto paper to map out the flow of the program, ultimately culminating in a "results" screen that offered the insight into the emails analyzed.

wireframe
Sketches
wireframe
Sketches
wireframe
Sketches

Deliver

I hired a CAD professional on Upwork to make the final renderings and animation shown below. I used these renderings to advertise the product to companies in cold calls and in emails.