Portrait of Daniel Short

Data Science

Data Science,
Made Actionable

I focus on the modeling side of my portfolio: NLP, computer vision, generative models, evaluation tradeoffs, and the engineering work required to get demos deployed.

Python · PyTorch · NLP · AWS

Scroll for results

Applied Modeling With Measured Impact

The strongest examples here connect modeling work to measurable outcomes: faster data-quality workflows, broader tracking coverage, stronger anomaly detection, and NLP-guided website improvements.

  • 95% Cut delivery time by re-platforming R-based quality workflows into a one-click Python app.
  • 10x Expanded serial tracking coverage with decision-tree logic at Randall Reilly.
  • 98% Reached anomaly precision in data-quality detection workflows.
  • +14.13% Raised VisitGrandJunction.com pageviews per user after NLP-guided site analysis and fix recommendations.

Applied ML That Ships

The best examples here are not just model experiments. They show how retrieval, evaluation, deployment, and interface choices work together to make the output useful.

Project Focus
Delivered Value
NLP retrieval
Faster knowledge search with better relevance than keyword matching
Model evaluation
Higher trust in predictions, rankings, and recommended outputs
Computer vision
Real-time pattern recognition inside a usable browser demo
Deployment
Working tools instead of static screenshots or notebook claims

Work Experience

Recent roles with measurable impact across data quality, applied modeling, reporting automation, and broader business analysis.

Target

June 2022 – November 2023

Asset Protection Data Analyst

  • Designed dashboards that boosted theft reporting by 57.6% and increased prevention 180%.
  • Reduced inventory loss by 24% through analytics-driven investigations and tooling.
  • Presented actionable insights to non-technical stakeholders across multiple teams.

Randall Reilly

November 2023 – January 2024

AI Data Quality Analyst

  • Re-platformed R workflows into a one-click Python app, cutting delivery time by 95%.
  • Expanded serial tracking by 10x with decision trees and delivered 98% anomaly precision.
  • Deployed an autoencoder model that lifted overall data quality for downstream analytics.

Visit Grand Junction

February 2024 – Present

Business Analyst

  • Automated Industry Data Reports and stakeholder reporting, saving 200+ hours annually and cutting turnaround 99%.
  • Authored semimonthly council and stakeholder updates using tourism, lodging, and web-performance data.
  • Identified lodging cannibalization by reconciling tax records with occupancy, ADR, and RevPAR, shaping council updates and strategy.
  • Researched traveler trends and AI search behavior to guide content and GEO strategy, contributing to +9.4% organic sessions and 10x AI referrals YoY.

Certifications & Degrees

Credentials and formal education supporting analytics and machine learning work.