About Me

I am a passionate data scientist who believes in data-driven decision making and empowering executives with actionable facts. I focus on bridging the gap between data science and data engineering, utilizing modern design patterns for cloud-based and containerized workflows to maximize data science impact. I currently work for CBS Interactive in the Applied Machine Learning Group

News

Aug 2018 I will be presenting "Serverless Streaming Web Analytics with AWS Kinesis and Lambda" at the Nashville Analytics Summit.
Mar 2018 I presented "Building a Serverless Recommendation Engine" at the Brentwood Artificial Intelligence meetup discussing how to use AWS Lambda and PySpark on AWS ElasticMapReduce using collaborative filtering to generate personalized recommendations for millions of users for just a few dollars.
Feb 2018 I attended PyTennessee 2018 in Nashville, Tennessee
Nov 2017 I attended AWS re:Invent 2017 in Las Vegas, Nevada
Jul 2017 I presented my work integrating our custom analytics with Splunk into our chatbot for on-demand analytics queries by our publishers at the Splunk Nashville Users Group.
Apr 2017 I began working at 247Sports.com at CBS Interactive
Mar 2017 I successfully defended my PhD in Electrical Engineering from Vanderbilt Univeristy entitled "Optic Nerve Characterization using Magnetic Resonance Imaging: The Search for Biomarkers" under the supervision of Dr. Bennett Landman (Electrical Engineering) and Dr. Seth Smith (Radiology and Biomedical Engineering).
Feb 2017 I attended SPIE Medical Imaging in Orlando, FL

Work Experience

Jul 2018 - Present CBS Interactive - Applied ML Group, Senior Machine Learning Engineer
  • Recommendation algorithm development and deployment to Kubernetes and GCP

Apr 2017 - Jul 2018 CBS Interactive - 247Sports.com, Data Engineer
  • Created serverless personalized recommendation engine using PySpark to get recommendations for >2 million users for a few dollars

  • Seamlessly moved analytics from custom PHP+MySQL application to AWS Kinesis streaming while maintaining business reporting

  • Created machine learning spam detection algorithm using Naive Bayes TF-IDF classifier, deployed to production

  • Created a serverless sessionization pipeline using AWS Lambda and AWS Elasticache (redis) to meet business needs of reporting session-level traffic metrics

  • Wrote chatbot to interface with analytics giving publishers real-time query insights into their performance metrics

  • Created custom serverless ETL pipeline of production data into BigQuery Data Warehouse for analysis.

Aug 2014 - Mar 2017 Vanderbilt University Department of Electrical Engineering and Computer Science, Graduate Research Assistant
  • Development of novel fully automated optic nerve segmentation, measurement algorithms in MATLAB

  • Deployment of algorithm using open source Python cluster manager (DAX) to process >11,000 clinical and research image volumes across self-maintained compute grid and university supercomputing resources

Projects

Personal Blog and Website https://unsupervisedpandas.com
  • Hosted on GitHub Pages. Uses Markdown for posts and Jekyll for static HTML generation.

  • Uses AWS Route53 for custom DNS resolution as well as Cloudflare CDN

  • Includes posts on data science and data engineering top tags include scikit-learn, AWS, Python, Apex, Docker, Lambda, Pandas and Data Engineering.

GitHub Portfolio http://github.com/Raab70
  • Hosts code samples, original source code, and patches for open source projects.

  • Open source contributor to multiple projects including matplotlib

NSS Partnership http://nashvillesoftwareschool.com/programs/data-science
  • Designed project for NSS Data Science track

  • Provided data and AWS cloud setup for students

  • Instructed students on working with big data in the cloud using PySpark on AWS EMR.

Skills

Languages

Python, Bash, terraform, SQL

Frameworks

Flask, NumPy, Pandas, PySpark, matplotlib, scikit-learn, TensorFlow

Technologies

Kubernetes, Containerization, AWS, Google Cloud, microservice architecture

Education

Aug 2013 - Mar 2017 Ph.D. in Electrical Engineering
Vanderbilt University
Aug 2011 - May 2013 Master's in Business Administration
The University of Tennessee at Martin
Aug 2007 - Feb 2011 B.S. in Imaging Science
Rochester Institute of Technology

Selected Publications

Google Scholar

Optic Nerve Characterization using Magnetic Resonance Imaging: The Search for Biomarkers
R. Harrigan
Vanderbilt University ETD 2017
[1]
Quantitative characterization of optic nerve atrophy in patients with multiple sclerosis
R. Harrigan, A. Smith, B. Lyttle, B. Box, B. Landman, F. Bagnato, S. Pawate, and S. Smith
MSJ-ETC 2017
[2]
Improved automatic optic nerve radius estimation from high resolution MRI
R. Harrigan, A. Smith, L. Mawn, S. Smith, and B. Landman
SPIE MI 2017
[3]
Disambiguating the optic nerve from the surrounding cerebrospinal fluid: Application to MS-related atrophy
R. Harrigan, A. Plassard, F. Bryan, G. Caires, L. Mawn, L. Dethrage, S. Pawate, R. Galloway, S. Smith, and B. Landman
MRM 2016
[4]

Honors & Awards

2014 - 2017 NEI T32 Training Grant Award
2015 ISMRM Trainee Educational Stipend Award Winner
2012 and 2013 UT Martin Faculty Scholar
2011 RIT Outstanding Undergraduate Scholar Nominee
2010 - 2011 RIT Academic Achievement Scholarship

Teaching Experience

Fall 2015 Vanderbilt University (VU Athletics), Tutor
Spring 2014 Digital Signal Processing (EECE256), TA
Spring 2013 Image Processing (EECE253), TA
Spring 2008 - Spring 2011 RIT Academic Support Center (Drop in calculus and physics tutoring), Lead Tutor

All Publications

Google Scholar

Journal Articles

Robust Multicontrast MRI Spleen Segmentation for Splenomegaly Using Multi-Atlas Segmentation
Y. Huo, J. Liu, Z. Xu, R. Harrigan, A. Assad, R. Abramson, and B. Landman
IEEE TBE 2018
[J1]
Quantitative characterization of optic nerve atrophy in patients with multiple sclerosis
R. Harrigan, A. Smith, B. Lyttle, B. Box, B. Landman, F. Bagnato, S. Pawate, and S. Smith
MSJ-ETC 2017
[J2]
Estimated Incidence of Ophthalmic Conditions Associated with Optic Nerve Disease in Middle Tennessee
S. Chaganti, K. Nelson, R. Harrigan, K. Nabar, N. Nandakumar, T. Goecks, S. Smith, B. Landman, and L. Mawn
2017
[J3]
Disambiguating the optic nerve from the surrounding cerebrospinal fluid: Application to MS-related atrophy
R. Harrigan, A. Plassard, F. Bryan, G. Caires, L. Mawn, L. Dethrage, S. Pawate, R. Galloway, S. Smith, and B. Landman
ISMRM 2016
[J4]
Automatic Measurement of Optic Nerve Atrophy in Multiple Sclerosis Patients with and without Optic Neuritis in MRI (P3. 013)
R. Harrigan, B. Landman, L. Mawn, S. Pawate, and S. Smith
AAN 2016
[J5]
Vanderbilt University Institute of Imaging Science Center for Computational Imaging XNAT: A multimodal data archive and processing environment
R. Harrigan, B. Yvernault, B. Boyd, S. Damon, K. Gibney, B. Conrad, N. Phillips, B. Rogers, Y. Gao, and B. Landman
Neuroimage 2016
[J6]
Robust optic nerve segmentation on clinically acquired computed tomography
R. Harrigan, S. Panda, A. Asman, K. Nelson, S. Chaganti, M. DeLisi, B. Yvernault, S. Smith, R. Galloway, and L. Mawn
SPIE MI 2014
[J7]

Conference Proceedings

Improved automatic optic nerve radius estimation from high resolution MRI
R. Harrigan, A. Smith, L. Mawn, S. Smith, and B. Landman
Orlando, Florida 2017
[C1]
Multi-atlas segmentation enables robust multi-contrast MRI spleen segmentation for splenomegaly
Y. Huo, J. Liu, Z. Xu, R. Harrigan, A. Assad, R. Abramson, and B. Landman
Orlando, Florida 2017
[C2]
Effects of b-value and number of gradient directions on diffusion MRI measures obtained with Q-ball imaging
K. Schilling, V. Nath, J. Blaber, R. Harrigan, Z. Ding, A. Anderson, and B. Landman
Orlando, Florida 2017
[C3]
Short term reproducibility of a high contrast 3-D isotropic optic nerve imaging sequence in healthy controls
R. Harrigan, A. Smith, L. Mawn, S. Smith, and B. Landman
San Diego, California 2016
[C4]
Structural functional associations of the orbit in thyroid eye disease: Kalman filters to track extraocular rectal muscles
S. Chaganti, K. Nelson, K. Mundy, Y. Luo, R. Harrigan, S. Damon, D. Fabbri, L. Mawn, and B. Landman
San Diego, California 2016
[C5]
On the fallacy of quantitative segmentation for T1-weighted MRI
A. Plassard, R. Harrigan, A. Newton, S. Rane, S. Pallavaram, P. D'Haese, B. Dawant, D. Claassen, and B. Landman
San Diego, California 2016
[C6]
Constructing a statistical atlas of the radii of the optic nerve and cerebrospinal fluid sheath in young healthy adults
R. Harrigan, A. Plassard, L. Mawn, R. Galloway, S. Smith, and B. Landman
Orlando, Florida 2015
[C7]
TRADITIONAL MBA ADMISSIONS CRITERIA AND GRADUATE SCHOOL SUCCESS: HOW ARE GMAT SCORES, UNDERGRADUATE GPA, AND GRADUATE BUSINESS SCHOOL PERFORMANCE CORRELATED?
K. Hammond, M. Cook-Wallace, E. Moser, and R. Harrigan
AELJ 2015
[C8]

Posters

Automatic Measurement of Optic Nerve Atrophy in Multiple Sclerosis Patients with and without Optic Neuritis in MRI
R. Harrigan, S. Smith, and B. Landman
Vancouver, BC, Canada 2016
[S1]
Imaging in the Optic Nerve: Insights into Optic Nerve Pathology
R. Harrigan, S. Chaganti, S. Smith, and B. Landman
Nashville, Tennessee 2015
[S2]
Optic Nerve Mapping in Multiple Sclerosis Patients with and without Optic Neuritis
R. Harrigan, S. Smith, and B. Landman
Nashville, Tennessee 2015
[S3]
Mapping of the Optic Nerve in MS Patients with and without Optic Neuritis
R. Harrigan, S. Smith, and B. Landman
Toronto, ON, Canada 2015
[S4]