Philippe Dolimier

Microsoft, Azure (Software Engineer 1)

2020 - present

I am currently operating, maintaining, and improving the Azure Windows operating system update deployment infrastructure. We deploy OS upgrades to ~1.65 million internal customers (e.g. Bing, Cosmos, and Xbox machines) and to ~3 million+ assets within the Azure public and private clouds, monthly. Most of my development is in C#/.Net. Additional tools used are ADO, Azure Data explorer (KQL). I have also recently implemented a scalable analytical engine to automate our high stakes validation testing for OS safety which makes the decision whether to release our updates to millions of customer workloads. (C#, Python, KQL, & various Microsoft internal tools)

Northrop Grumman, Missions Systems (Associate Software Engineer)

2017 - 2020

2019: Supported customer workstation development projects via Open Source application modification, creating kernel extensions, and implementing various security plug-ins. Our team's aim was adding security to mac OS client machines.

2018: Focused on implementing simulation of an EO/IR sensor processing device which receives data from RADAR systems. Conducted research regarding localizing text from an unstructured scene, using some MSER variations.

2017: Produced workflow processes for Management to streamline redeployment and training procedures. Created client-side in basic HTML, CSS, and JavaScript. Using Ajax Requests and Server-side PHP to access and store employee data from server.

University of Maryland Teaching Assistant (Data Analytics)

Spring 2019

Hosted office hours for students to ask questions and garner answers regarding class material, homeworks, and projects.
Graded projects and homework assignments.
Helped to fill gaps in students' knowledge conceptually and technologically.

Engineering ASB Trip Coordinator

Spring 2017, Spring 2018

University of Maryland's Engineering Departments Habitat for Humanity trip coordinator, performing community service, building homes for the underprivileged in Pinehurst, North Carolina.
Co-ordination tasks included recruiting volunteers, communication with the Pinehurst Habitat for Humanity local chapter, budgeting funds, and acting as a liaison between the University and Habitat.

Energy Efficient Technologies Intern

October '14 - January '15

Created excel spreadsheets and graphs to find discrepancies in energy use.
Built energy models and marketing tactics for a HVAC systems monitoring product called 'exybox'.
Learned to code in Python and HighCharts to showcase data for HVAC efficiency.

Romilos Restaurant

June '13 – August '16

My first real job experience. I began as a Host and Busboy, then became a Food Server and worked in Food Preparation.
I eventually helped to create and implement weekly staff work schedules. Working here throughout high school 3-4 days a week, I developed a strong work ethic.

The Fad Name Mystique

April '18

Some of my most prized work, this project outlines a full-scope and formal exploration of name fads and trends from 1910 - 2016. I begin with extensive Data Curation, Exploratory Data Analysis, and Hypothesis Testing using data from the Social Security Administration's immense baby names repertoire. Follow along in my Rmarkdown project file as I attempt to answer complex questions about the relationship between gender, time, and names through a complex 'trend metric' of my conception.

View Project »

Plus Interest

March '18

In my Spring 2018 semester I created this project as a part of my Node.js Development Course. It is a peer to peer lending platform using Handlebars templating, Node.js, and MongoDB. Currently Hosted using 'now' deployment services. I have a few ideas to progress the app, one being the transition to a charity service for individuals in developing countries to fund their small businesses, another being a fully-fledged lending platform with completely customizable loans, made profitable by means of priority lending.

View Project »

Dating Site

March '18

In my Spring 2018 semester I created this project as a part of my databases course. This was the first time I learned how to make a Full-Stack site using best practices. I first designed my database schema to be a users table with the match relationships as a second table dependant of the first. I then created the back-end logic in Java, and used Tomcat to run my servers locally. This was all in conjunction with a variation of HTML (JSP) to create webpages (front-end). The site features the ability to make an account and find your best matches based on hobbies, major, town, and ratings.

freshDressed

December '16 - January '17

This project is meaningful to me since it is my first solo project with no outside direction. I took a problem I had and worked to make it from scratch, researching heavily along the way. I Founded and Coded 'freshDressed', which is an interactive weather application portal. It uses web-scraping scripts to pull in weather data from a weather site. It then computes an optimal outfit for any day in any national Zip Code taking into account the entire days temperatures, humidity, chances of precipitation and wind, implemented in Java. It is then viewed via the web on an interactive, aesthetically pleasing platform written in Web Development languages.

The Repetitiveness of Pop Music

Summer '18 (Current Project)

This is a project currently in the works. I am studying the largely contested question, Has Pop Music become more repetitive over time? Using Kaggle Data sets, I am aiming to compress data to achieve a repetition metric based upon the compressibility of a file (LZ-compression). The more compressable a file, the more repetitive it is. I intend to next begin pulling data from Spotify's API to blend the data set for further applications and explorations of the new metric.

View Project »

AirBnB Data Set: Prediction Contest

October '18 - December '18

This Writing Sample culminates the semester-long project with the goal of predicting which properties from a hidden data set have a high booking rate (booked > 80% of the year), using a Kaggle Training data set with 40,000 entities and around 40 variables.
Our team had the highest prediction accuracy of any group at 85.53%, a 10.43% increase from the 75% baseline.

We believe much of our success can be contributed to our extensive data cleaning process, including the way we encoded missing data, the way we partitioned variables into sub-category variables, and to our diversity of models created, ultimately garnering our best prediction accuracy with our Random Forest Ensemble method.

Credit Cards' Impact on Consumer History

May '18

An explanatory and persuasive essay on the impacts of the credit cards on consumer history through a historical lens, exploring cultural changes through the concepts of consumers' detachment between cash and card, the process of cultural acceptance of debt, psychology used in advertisements for credit cards, and economist's viewpoints of the credit boom.

Department of Motor Vehicles Efficiency Research Project

September '17 - December '17

This final Paper is the culmination of a semester long research project on the Maryland DMV (MVA). I chose to conquer a widely accepted issue, the inefficiencies that lie in the DMV. I set a constraint on myself that I would complete this hypothetical project under the strict budget that the sector has.

Contemporary Living-Dead Legends Research Project

April '17

What I believe to be one of my most eloquently written and elucidative essays, this project outlines the research conducted in a course at the University of Maryland about societies' fears and how they affect our culture. This being how a society's stories and legends about the living dead stem specifically from their respective practices with their dead, theories about the afterlife, and the common fears of their societies, as the current events of a time lead to the myths and stories embedded in that specific culture.