Matthew Merrill

Matthew Merrill

Data Scientist

Fellowship.ai

About Me

I specialize in helping others envision the complexities of data so progress can be made.

As a data scientist with Fellowship.ai I worked on a range of projects, from CTOR prediction for email banners to customer segmentation analysis and churn prediction. I’ve also deployed workable models into Streamlit apps, R Shiny dashboards and Docker containers for exploration and practical use. Beyond my programming work, what sets me apart is my love for telling compelling data stories that bring clarity and drive decisions.

Links to my personal projects are below, feel free to reach out to me with employment opportunities, freelance work or to pair up on an interesting project.

Interests

  • Machine Learning for Good
  • Rethinking Education with AI
  • Compelling Data Stories

Education

  • Data Science Career Track Certificate, 2020

    Springboard

  • Masters, Mathematics Education, 2015

    University of San Francisco

  • Bachelor Physics, minor in Math and Astrophysics, 2012

    University of San Francisco

Python Programmer

Problem Solver

Data Story Teller

Projects

Links to code and report repositories in github.

*
All Machine Learning Classification Recommendation Systems User Insights Capstone Projects Mini-Projects Data Visualization Deep Learning NLP

Predicting Churn

Implementation of deep learning and Causal ML approaches to predict customer churn on dummy dataset.

Sentiment Analysis with ULMFiT

Project aimed at differentiating between positive and negative reviews using the fastai’s ULMFiT implementation method.

Streamlit App

Interact with the data from two hotels in Portugal to see the affect of altering customer markets and lead time on cancellation rates.

Mini_projects

A repository of projects completed through the Springboard career track program.

Predicting Rider Retention

Predicted rider retention for a taxi service and identified most significant factors that contributed to it. Achieved an 80% accuracy with a catboost model, which was chosen for its interpretability.

Predicting User Adoption

Identified the most important factors contributing to user adoption for a product. Achieved a 96% accuracy with an SVM model, and found user length and opting into the mailing list as the most significant predictors.

Lightfm Recommendation System

Recommendation engine with a .97 AUC achieved using clustering techniques to create user features. Data represents Olist marketplace transactions and was retrieved from kaggle.com.

Predicting Hotel Cancellations

Created prediction algorithm for determining if a customer will cancel at the moment of booking. After eliminating numerous data leakage sources, we achieved a 90% AUROC with a catboost classification model.

Experience

Data centered achievements.

 
 
 
 
 

Machine Learning Fellow

Fellowship.ai

Jan 2021 – Apr 2021 San Francisco, California

Led as scrum master for a team of 10 data scientists, spearheading projects in CTOR optimization for email campaigns, churn prediction and age segmentation.

  • Implemented agile software methods to solve problems in campaign image optimization, churn prediction and customer segmentation for Dockers.
  • Lead as scrum master for a team of 10 data scientists, reporting daily progress and presenting weekly slide decks to the founder.
  • Improved CTOR prediction accuracy by over 30% compared to baseline models and customer age prediction by 20% using oversampling methods.
 
 
 
 
 

Data Science Student

Springboard

Jan 2020 – Dec 2020 San Francisco, California

Completed 600+ hours of hands-on curriculum, with 1:1 industry expert mentor oversight, and completion of 2 in-depth projects.

  • Skills Learned: Python, R, SQL, Machine Learning (supervised and unsupervised), Data Wrangling, Exploratory Data Analysis, Feature Engineering, Data Visualization, Predictive Modeling, Time Series Analysis, Classification, Regression and Clustering.
 
 
 
 
 

Math Department Chair

SFUSD

Aug 2015 – Jul 2020 San Francisco, California

Taught with a data focused approach to closing the achievement gap.

  • Created a data-driven culture among departments to influence changes to school-wide pedagogy.
  • Led a cross-functional initiative to increase participation in tutoring programs, leading to an 11% increase in assessment scores.
  • Conducted 4 semesters of school wide PD sessions with a team of colleagues, communicating monthly results of research and experimental data collection.

Accomplishments

More to come!

Data Science Career Track

Springboard Jan 2020 – Sep 2020
600+ hours of hands-on course material, with 1:1 industry expert mentor oversight, and completion of 2 in-depth portfolio projects.

Popular Topics

Select a tag to see related projects.

Business Insight Capstone Project Catboost Classification Clustering Data Leakage Data Science Data Visualization Deep Learning Experimental Design Feature Engineering Feature Importance Inferential Statistics LightFM Machine Learning Mini-Project Mini-Projects NLP Time-Series Analysis User Insights

Contact

Feel free to reach out for freelance work, to co-author a project, or inquire about my experience.