Colby Dolly, Ph.D. · Director of Science and Innovation

QUALITY

Lean Six SigmaBlack Belt Certified

LEADERSHIP

10+ YearsAI & Analytics Strategy

RESEARCH

Ph.D.Political Science · View Study →

GOVERNANCE

Prosci®Change Practitioner

About

I am an accomplished AI and analytics leader with over a decade of experience driving enterprise-wide technology adoption, advanced analytics, and evidence-based innovation across research, healthcare, and public sectors. Currently, I serve as Director of Science and Innovation at the National Policing Institute (NPI) in Arlington, VA.

In my current role, I lead the enterprise deployment of large language models (LLMs) and generative AI, establish organizational AI governance frameworks, and architect custom Retrieval-Augmented Generation (RAG) and agentic workflow systems. I also serve as Principal Investigator on federally funded research evaluating computer vision and NLP models for operational public safety applications.

Prior to joining NPI, I led enterprise analytics and real-time KPI data pipeline infrastructure at Centene Corporation and founded the Crime Analysis Unit for the St. Louis County Police Department. I hold a Ph.D. in Political Science from the University of Missouri–St. Louis specializing in organizational change and quantitative methodology, an M.B.A. from Lindenwood University, and am a Prosci® Certified Change Practitioner. Over the course of my career, I have secured over $3.5 million in competitive research awards connecting advanced machine learning with measurable organizational impact.

Skills & Expertise

Generative AI & LLMs: Anthropic API (Claude)
Claude Code
RAG Architecture
Vector Databases
Agentic Workflows
Custom Skills & MCP
Prompt Engineering
AI Strategy & Governance: Enterprise AI Adoption
AI Acceptable Use Policy
Risk & Compliance Guardrails
Change Management (Prosci®)
Workforce AI Enablement
Data Science & ML: Natural Language Processing (NLP)
Predictive Modeling
Multivariate Regression
Logistic Regression
ANOVA & Factor Analysis
Programming & Tools: Python (Pandas, NumPy)
R (Tidyverse, Shiny, ggplot2)
SQL
Git / GitHub
Bash & Automation
LaTeX
Spatial & GIS: ArcGIS Pro
ArcGIS Online
QGIS
Spatial Analytics & Geoprocessing
Research & Evaluation: Randomized Controlled Trials (RCT)
Quasi-Experimental Design
Survey Design & Mixed Methods
Grant Proposal Architecture

Selected Publications & Reports

Improving Police Encounters in Pedestrian Stops: The Phoenix Quasi-Experiment

Colby Dolly, David Weisburd, M. Valdovinos Olson, B. Dong

National Policing Institute, (Award No. 2019-R2-CX-0025), 2026

BibTeX

Citation · BibTeX

@techreport{dolly2026improving,
  title       = {Improving Police Encounters in Pedestrian Stops: The Phoenix Quasi-Experiment},
  author      = {Dolly, Colby and Weisburd, David and Valdovinos Olson, M. and Dong, B.},
  institution = {National Policing Institute},
  number      = {Award No. 2019-R2-CX-0025},
  year        = {2026},
}

Considerations for Specialized Units: A Guide for State and Local Law Enforcement Agencies to Ensure Appropriateness, Effectiveness, and Accountability

Colby Dolly, J. Cherkauskas, E. Hansen, E. Renouard

Office of Community Oriented Policing Services (COPS), 2024

BibTeX

Citation · BibTeX

@techreport{dolly2024specialized,
  title       = {Considerations for Specialized Units: A Guide for State and Local Law Enforcement Agencies to Ensure Appropriateness, Effectiveness, and Accountability},
  author      = {Dolly, Colby and Cherkauskas, J. and Hansen, E. and Renouard, E.},
  institution = {Office of Community Oriented Policing Services (COPS)},
  address     = {Washington, DC},
  year        = {2024},
}

Multi-modal Analysis of Body-Worn Camera Recordings: Evaluating Novel Methods for Measuring Police Implementation of Procedural Justice

Colby Dolly, E. Hansen, R. Gallardo, J. Coleman, J. Wender, B. Lande, V. Ta-Johnson, M. Tomlinson, G. Ben-Yosef, P. Tu

National Policing Institute, (Final Project Report No. 2020-R2-CX-0010), 2024

BibTeX

Citation · BibTeX

@techreport{dolly2024multimodal,
  title       = {Multi-modal Analysis of Body-Worn Camera Recordings: Evaluating Novel Methods for Measuring Police Implementation of Procedural Justice},
  author      = {Dolly, Colby and Hansen, E. and Gallardo, R. and Coleman, J. and Wender, J. and Lande, B. and Ta-Johnson, V. and Tomlinson, M. and Ben-Yosef, G. and Tu, P.},
  institution = {National Policing Institute},
  number      = {Final Project Report No. 2020-R2-CX-0010},
  year        = {2024},
}

View all publications & reports →

Featured Projects

Nationwide Public Safety Confidence Index & Dashboard

Inaugural nationwide dashboard launched with Zencity and NPI to continuously track public sentiment, community trust, and safety metrics.

Multi-Modal Analysis of Body-Worn Camera Recordings

Evaluating computer vision and NLP models to measure police implementation of procedural justice under NIJ Award 2020-R2-CX-0010.

Enterprise RAG & Agentic Workflows for SOW Automation

Architected internal RAG vector search and Claude-powered agentic pipelines, reducing proposal authoring turnaround time by 80%.

View all projects →