The who, what, why, and how of this project

Danny Morris 2021-02-19 2 min read

Who we are

The Data Science Squad is a diverse, global team of data and analytics professionals who organized for the experience of developing a production-grade software product powered by machine learning.

The Team:

Nisrine Hammout - Machine Learning Engineer

Gayatri Dobhal - Machine Learning Engineer

Eric Caskey - Data Engineer

Melvin Sebastian - Data Engineer

Cassin Thangam Edwin - Data Visualization Engineer/Front-end Developer

Abdul-Mateen Qamardeen - Website Developer and QA Analyst

Danny Morris - Team Lead

What we are building

We are building a full stack software product that uses machine learning to forecast short-term crime volume throughout the various geographical areas in the city of Buffalo. The idea is to support the Buffalo Police Department with a data-driven service that helps them schedule and deploy an appropriate number of patrol officers given the predicted crime volume in the coming days and weeks.

This software has 4 main components:

  • Database: Centralized data storage for all production data assets.

  • Machine learning model: Model for predicting future crime volume throughout the city of Buffalo.

  • Front-end web application: Dynamic UI for users to interact with production data assets.

  • Project website: A place to brand and promote our work.

Why we are doing this

We are working together to gain experience building a high-quality product, collaborate with like-minded professionals, and learn.

How are we doing this

To accomplish our goals, we must rely on two key values: communication and automation.


Use our Slack workspace to ask questions and issue updates as frequently as possible. Use our Trello board to lay out your current tasks and update their status often. Finally, get familiar with the required Git workflows(feature branching and pull requests) to “communicate” your work in the form of code.


When developing new features, always ask yourself “can this be automated?” Our final product will be set to update automatically each day, and for that to work all aspects of the product must be fully automated. To achieve automation we will make use of GitHub Actions, Streamlit Sharing, and possibly other automation tools if the need arises.