My Resume

Click each section below to expand resume.

Experience

University of North Carolina at Charlotte - Charlotte, NC
Data Scientist
June 2022 - Present

Work on the Charlotte Regional Data Trust where I securely integrate individual‐level administrative data across agencies in the Charlotte–Mecklenburg region. By linking records through a robust, privacy‐preserving system, the Data Trust is able to repurpose administrative data for research, evaluation, and planning.
Co-authored the inaugural State of Our Data report which highlights different statistics and visualizations about people in the Charlotte-Mecklenburg region that are in the data we maintain. Given that much of the data reflects marginalized or underserved communities, the report aims to equip changemakers with the tools, systems, and insights needed to address regional social, economic, and environmental challenges.
Developed an entity resolution process that links individuals across datasets using names, birthdates, and demographic information. Created with a synthetic dataset containing fake names, birthdates, addresses, gender, and race. Another dataset was created from this dataset that had variations and noise (hyphenated names, misspellings, twins) introduced to simulate real-world inconsistencies. The matching process involves five key steps: preprocessing (standardizing formats, removing noise, Soundex phonetic encoding), indexing to find possible matching candidates, comparing candidate pairs using string similarity and field weighting, classifying matches, deduplicating matches and assigning anonymized IDs.
Created an Airflow orchestration pipeline that processes data deposits from over 12 data partners, handling various formats and cleanliness levels by unzipping files, converting them to CSV, designing database schemas and loading the data into a database. It includes automated data cleaning based on field and deposit types, record comparison for database inserts or updates, and comprehensive logging for error tracking. The process will also generate stats for every field which feeds into visualization database tables used in Amazon Web Services (AWS) Quicksight business intelligence dashboards.
Built a web mapping application called Regional Explorer in Angular that leverages the ArcGIS JavaScript Application Programming Interface (API) and hosted feature layers from ArcGIS Online to visualize regional trends. The tool presents indicators across demographic, economic, and educational categories at the Census Tract level for a 14-county region supported by UNC Charlotte’s Urban Institute. Designed for use by community members, researchers, and decision-makers, it enables users to explore how areas are changing over time. Built-in analysis tools include correlation analysis between indicators and Moran’s I to identify spatial clustering patterns.

University of North Carolina at Charlotte - Charlotte, NC
GIS Developer
November 2017 - June 2022

Administer ArcGIS Server, Portal and NGINX servers hosting mapping applications to support Facilities Management operations.
Maintain the Campus’ web map application used by students, staff and visitors to find buildings and services on campus.
Develop geospatial solutions to help groups solve different problems such as creating a utility network to support better data management and setting up maps and database schemas for field data collection of assets.
Create interactive Tableau dashboards to highlight work request performance metrics for FM.

University of North Carolina at Charlotte - Charlotte, NC
Data Science Intern
August 2021 - December 2021

As part of my Masters of Data Science program I had an internship with the campus' Facilities Management department where I used machine learning to solve different problems.
XGBoost regression models to predict the time and cost for a work request to be completed.
Word2Vec and Deep Learning models to classify the type of problem from work request descriptions.
Analyzing the survival probability of different equipment categories over time.

New York City Department of Education - Brooklyn, NY
GIS Developer
April 2017 - August 2017

Created and maintained spatial datasets for the New York City school system which consisted of data for over 1,800 schools and 1.1 million students.
Maintained an ASP.NET mapping application and the related ArcGIS map services, and automated GIS tasks using Python.
Developed a prototype web mapping application as a Proof of Concept using only Open Source technology.

Mott MacDonald - Iselin, NJ
GIS Specialist
June 2007 - March 2017

Developed Python scripts for a variety of uses such as FTP image transfer, spatial analysis, creating map books and developing geoprocessing services.
Developed an ASP.NET mapping application to view and edit data which played a pivotal role in acquiring new clients.
Architected a document management system that linked documents to spatial assets in a web-based GIS application.
Administered a production GIS Server environment to host data for approximately 25 clients.

Rutgers University - Newark, NJ
Assistant Professor
January 2011 - May 2011

Taught a course on object-oriented programming covering concepts such as classes and inheritance.
This program was part of a certification for veterans of Iraq and Afghanistan that provided them tools to start a career in the Information Technology field.

Essex County College - Newark, NJ
Assistant Professor
February 2009 - May 2009

Taught a course that was part of a GIS certificate program covering analysis techniques using ArcGIS Desktop and the Spatial, Network and 3D Analyst extensions.

Union County NJ - Bureau of GIS - Elizabeth, NJ
GIS Technician
August 2005 - June 2007

Designed maps for different County departments ranging in uses from maps used in court trials to County park maps.
I also supported various County departments with assistance using the ArcGIS software suite.

Education

University of North Carolina at Charlotte, Charlotte, NC
Master of Science - Data Science and Business Analytics
September 2020 - December 2021

4.0 GPA

University of North Carolina at Charlotte, Charlotte, NC
Graduate Certificate - Data Science and Business Analytics
January 2019 - December 2019

4.0 GPA

Oregon State University, Corvallis, OR
Graduate Certificate - Geographic Information Science
January 2012 - May 2014

3.93 GPA

Kean University, Union, NJ
Bachelor of Science - Computer Science
September 2001 - May 2004

Projects

Finding Clusters of NYPD Complaints

Final Report

Used Neo4J to load NYPD complaint data and identify clusters of complaints using Louvain community detection and PageRank centrality to find the most influential locations that are central to other nearby areas where similar offenses occurred.
Created an interactive map that was integrated into Jupyter notebook using ipyleaflet to show where hotspots of compalaint clusters are occurring.
Used machine learning to perform graph link prediction to predict the locations where future offenses will occur.

Analyzing Twitter Data for Popeye's Chicken Sandwich Release

Final Report

Used R with the Twitter API to collect tweets about the 2019 Popeye's chicken sandwich release.
Developed an emoji cloud in Tableau to capture the range of emotions over time.
Created dashboards showing sentiment analysis, topic modeling, and the change in retweet volume over time.
Performed network analysis to study the centrality of tweets about the fights that occurred at Popeyes.

External Camera Mapping Project

Built a secure application in our GIS Enterprise Portal that allows the Campus police department to view the locations of their external cameras on Campus through a web based viewer and mobile device.
Built a separate application that allowed them to edit the camera locations and attributes and attach images to the camera point features.

ArcGIS Enterprise Upgrade

Implemented a complete enterprise upgrade of the campus' GIS technology.
Upgraded the GIS Server and Portal applications, migrated maps from ArcDesktop to ArcGIS Pro and configured security best practices by setting up secure services, firewall rules, reverse proxies and Shibboleth authentication.

Opioid Addiction Dashboard

Project Dashboard

Developed an R Shiny application that showcased data about Opioid addiction in the United States.
Application was interactive and allowed the users to filter by different states and select a range of years.
Used a word cloud, bar charts, geospatial maps and a radar chart as part of the dashboard.
This was a final project for a Visual Analytics course where my group won best project.

Pedestrian Crossing Model

Worked in Qatar with the Ministry of Transport to analyze a road network and assist with the development of a network model and data visualization tool to find suitable locations to build pedestrian crossings.
Analysis was related to an infrastructure improvement project to accommodate the high volume of people expected for the upcoming 2022 FIFA World Cup.

Open Source Technology Proof of Concept

Developed a prototype web mapping application as a Proof of Concept using Open Source technology as a contingency plan in the event the city's contract with their current GIS vendor did not get renewed.
Used GeoServer to connect to an SQL Server Spatial database to publish layers, OpenLayers to consume the OGC WMS and WFS services in a web application and Turf.JS to perform analysis operations.
Provided a detailed report where I published the technology, methods and findings.

Web Map Editing Tool

Developed a web mapping editing tool that allowed users to perform CRUD operations on spatial data stored on an enterprise Oracle geodatabase.
Used for updating data for a large pipeline construction project spanning over 100 miles.
Incorporated audit tracking and role based functionality in order to control and log changes in the application.

Pipeline Risk Model

Managed the creation of a geospatial risk model for a natural gas pipeline system spanning two states that determined which pipe segments in their system have the highest probability of failing.
I collaborated with subject matter experts to develop formulas to calculate the probabilities of failure.
Analyzed data for the calculations, ranging from a variety of factors such as pipe segment features, soil data analysis and field survey reports.

Nicholas Occhipinti

My Resume

Click each section below to expand resume.