Data Analyst AI & Automation Enterprise Systems

Turning complex data into reliable, human-centered systems.

I build data-driven tools and workflows that turn complex operational information into useful, reliable systems. My work spans Python, R, GIS, enterprise data quality, automation, and applied AI, with a focus on human-centered and accessible technology.

Python
R
SQL
GIS
Applied AI
Data Quality

View projects Download resume GitHub LinkedIn

120+ TBmedia data analyzed

68 locationsgeocoded with Python

For hiring teams

Applied analytics and operational impact

Explore automation, enterprise data quality, GIS, and workflow improvement grounded in real organizational needs.

View applied work For graduate programs

Technical preparation and research direction

Review applied AI, public-health research, responsible modeling, and preparation in programming, mathematics, and statistics.

View research profile

Selected work

Projects with evidence, not just descriptions.

Browse all projects

Reviewed intakeMetadata
WizardQC-ready
output

AI-Assisted Metadata Wizard

Human-in-the-loop desktop automation for media metadata, accessibility alt text, and controlled-vocabulary review.

Problem: Manual metadata entry was repetitive, while AI-generated descriptions still required professional review.
Solution: Combined reviewed intake parsing, XMP/IPTC writes, alt-text drafts, and keyword QC in one application.
Result: An operational, human-reviewed workflow supporting a documented summer backlog of 2,628 files.

Try demo Case study

Storage analysis output summarizing folders and file sizes

Enterprise Media Storage Analysis

A filesystem analysis pipeline built to make decentralized institutional storage measurable.

Problem: Storage records were distributed across directories and operational systems.
Solution: Scanned directories, reconciled job records, and produced structured analysis.
Scale: 302,933 folders and more than 120 TB reviewed.

Case study Report

ArcGIS Pro map showing geocoded Hope Supply Co. partner locations in North Texas

Hope Supply Geocoding Pipeline

An interactive Python workflow that converts incomplete organization records into mapping-ready spatial data.

Problem: Partner records lacked full addresses and coordinates.
Solution: Automated ingestion, geocoding, field creation, and low-confidence review.
Result: 68 partner locations prepared for mapping.

Case study GitHub

Line chart showing U.S. maternal mortality rates across 12-month-ending estimates

U.S. Maternal Mortality Trends & Demographic Disparities

A post-class revision that turns a maternal-health research paper and supporting ML experiment into a defensible descriptive surveillance analysis.

Question: What trends and demographic differences can the CDC/NCHS snapshot support?
Approach: Quality flags, separate demographic analyses, rolling-window visualization, and June endpoint comparisons.
Revision: Target leakage identified, predictive claims removed, and AI assistance disclosed.

Case study Research paper

Bar chart showing outcomes for TSA claims

TSA Claims Analysis with R

Exploratory analysis of U.S. airport claims, their locations, outcomes, and highly skewed values.

Problem: Messy public claims data made national patterns difficult to interpret.
Solution: Cleaned, summarized, and visualized claim types, sites, amounts, and outcomes.
Finding: 24.4% of claims were approved across 94,848 records.

Case study Report

Professional context

Technical work shaped by real operations.

My background spans academic data operations, institutional media systems, automation, and client-facing production. That context keeps my technical work grounded in quality, documentation, accessibility, and the people who use the system.

View experience

Let’s connect

Interested in thoughtful data work?

I am open to data analysis, automation, enterprise-systems, GIS, and applied-AI opportunities, as well as graduate research conversations.

Email Crystal View resume