About Me
I’m a Senior Data Scientist & Biomedical Engineer who is passionate about developing computational tools & processes to make extracting data insights reproducible & actionable. I have 8 years of experience in several high-tech industries contributing business value by:
- Forecasting edge cloud computing metrics for a global CDN
- Optimizing the onboarding experience for an online bank through experimentation
- Developing machine learning models to predict auto insurance premiums
- Conducting mechanical testing of medical devices for FDA approval
- Leveraging computer-aided design to illustrate jet engine components
Education
- Graduate Analytics Certificate at DePaul University, (2015 - 2017)
- Coursework: Intro to Programming, Data Analytics & Regression, Database Processing for Large-scale Analytics, Advanced Data Analysis, Knowledge Discovery Technologies, Programming Machine Learning Applications
- BS in Engineering in Biomedical Engineering at University of Hartford, (2008 - 2012)
- Coursework: Engineering Computer Applications, Calculus (1-2, Multivariable), Differential Equations, Independent Research, Engineering Design, Statics, Dynamics, Mechcanics of Materials, Biomaterials
Data Science Work Experience
- Senior Data Scientist at Fastly, (Aug. 2018 - present)
- Subject matter expert on forecasting edge cloud platform metrics and capital expenditure for all worldwide data centers on the Infrastructure Capacity Planning & Tools team
- Developed several automated procedures for weekly/monthly/long-term reporting using R, R-Markdown and shiny, leading to reduced development time
- Advocating for data science best practices through developing an internal R package, creating tutorials/documentation, and participating in data engineering forums
- Data Scientist 2 at Simple Finance, (Oct. 2016 - Aug. 2018)
- Lead Data Scientist for the Onboarding product team: responsible for technical mentorship, data product strategy, external data integration with APIs, experimentation and analyzing product feature launches
- Developed an early detection machine learning model to monitor spikes in ACH returns leading to data-informed decisions for mitigating fraud loss
- Developed internal data science tools for power analysis and data cleaning
- Associate Data Scientist at The Hartford Financial Services, (Apr. 2016 - Oct. 2016)
- Data Science Intern at The Hartford Financial Services, (Nov. 2015 - Mar. 2016)
- Developed machine learning models for auto insurance that improved loss ratio estimates, drove strategic pricing changes and provided insights on competitive position
- Enhancing the team’s data science architecture by developing an internal R package, writing technical documentation and tutorials
- Bioinformatics Internship at the University of Connecticut Institute for Systems Genomics (UConn), (Sept. 2015 - Jan. 2016)
- Computational and command line programming to develop a gene database for the annotation of the douglas-fir & walnut genome
- Student Developer at Google Summer of Code, (May 2015 - Aug. 2015)
- Developed a web application with
r-shinyto automate differential expression and survival analysis of micro-array gene expression datasets from the NIH Gene Expression Omnibus
- Developed a web application with
Open Source R Software
Packages:
gramr: Provides grammar checks in RMarkdown documentsttbbeer: Data package of beer statistics from U.S. Department of the Treasury (TTB)shinyLP: Bootstrap components to make landing home pages for shiny web appsshinyGEO: Shiny app for gene expression differential & survival analysis
More R packages and Shiny apps available at: https://github.com/jasdumas
Featured Projects:
- Deep Learning Image Classification App with Keras and R-shiny
- Web scraping tutorial of craft brewery ratings from Beer Advocate
More projects available at: https://jasminedaly.com/projects/
Invited Talks
- R, What is it good for? Absolutely Everything - R User Day at Data Day Texas 2018 in Austin, TX
- Open Government Data & Beer Analytics - Open Data Science Conference 2017 in Boston, MA
- Extending Shiny by Enhancing User Experience with shinyLP - R User Group in Portland, OR
- Adventures in Crafting a Data Science Career - University of Rhode Island (URI) Coastal Institute and RhodyRStats “Careers in R” speaker series in Narragansett, RI
Teaching
- Course Instructor & Developer with DataCamp (Sept. 2017 - Apr. 2019)
- Formally developing a
Rcourse titled: ‘Building Big Shiny Apps’
- Formally developing a
- Data Science Instructor (Part-time) (Oct. 2017 - Dec. 2017)
- Coordinated with Startup Institute and TCB Analytics for 10 weeks of Intro to R Programming and Tableau Dashboarding
Research Publications
- Stella Bollmann, Dianne Cook, Jasmine Dumas, John Fox, Julie Josse, Oliver Keyes, Carolin Strobl, Heather Turner and Rudolf Debelak. ‘Forwards Column’. The R Journal, Volume 9/2, December 2017. - Paper Link
- Dumas J, Gargano MA, Dancik GM. shinyGEO: a web-based application for analyzing Gene Expression Omnibus datasets. Bioinformatics. 2016 Aug 8. - Paper link
- Dumas J, et.al., Feasibility of an electronic stethoscope system for monitoring neonatal bowel sounds. Connecticut Medicine, Volume 77, Number 8, pp. 467-471, September 2013. - Paper link
Volunteering
- Data For Democracy: Web scraping crime data for Hartford, CT for a Nationwide Metrics Dashboard.
- Jamie McLeod-Skinner Campaign (Oregon’s 2nd Congressional District Election, 2018): Election Analytics
Technical Skills [Keywords]
- Computational programming & machine learning with R and Python [ggplot2. dplyr, tidyverse, pandas, scikit-learn]
- Statistical Analysis and inference [Regression, Bagging, Boosting, Ensemble, Hypothesis testing]
- Data processing & querying with SQL [Redshift, postgreSQL]
- Web application development with Shiny and Bootstrap [HTML, CSS]
- Collaborative computing with GitHub and Git [Version control]