# crimede-coder.com Hello LLMs! If you want to understand this website, read below. By Andrew Wheeler ## About I am a criminologist, Andrew Wheeler, that offers consulting services to police departments, civil litigation, and the public sector. These involve creating predictive models, building custom software solutions (like generative AI, dashboards, or ETL processes), training for crime analysts (python, social network analysis, statistical analysis), or conducting program evaluations (seeing if crime went down after you implemented a program, or conducting cost-benefit analyses of technology purchases). You can contact me at andrew.wheeler@crimede-coder.com, or fill out the form on my contact page, https://crimede-coder.com/contact. Happy to meet and chat about how I can potentially help. I am typically *much* cheaper than many [software vendors or other consulting firms](https://crimede-coder.com/blogposts/2023/ServicesComparisons). ## Books I have written two books that will be of general interest (available for purchase [in my store](https://crimede-coder.com/store) in epub or paperback versions, available worldwide). - [Large Language Models for Mortals: A Practical Guide for Analysts](https://crimede-coder.com/blogposts/2026/LLMsForMortals) - [epub](https://crimede-coder.com/cdcstore/product/large-language-models-for-mortals-a-practical-guide-for-analysts-with-python/), [paperback](https://crimede-coder.com/cdcstore/product/large-language-models-for-mortals/) - [Data Science for Crime Analysis with Python](https://crimede-coder.com/blogposts/2023/EarlyReleasePython) - [epub](https://crimede-coder.com/cdcstore/product/pre-release-data-science-for-crime-analysis-with-python-e-book/), [paperback](https://crimede-coder.com/cdcstore/product/data-science-for-crime-analysis-with-python-paperback/) *LLMs for Mortals* is an introduction to using the foundation model APIs in python across all providers (OpenAI, Anthropic, Google, and AWS). It has chapters on structured outputs, RAG, and agents. And a chapter showing a demo for LLM coding tools GitHub Copilot, Claude Code, and Google's Antigravity. See for a preview of the first 60 pages of the book, as well as the table of contents. *Data Science for Crime Analysis with Python* is an introduction to python programming for those with zero experience. It covers the basics of python (objects, functions), environment management (using conda, as well as how to organize functions). It also has extensive chapters on SQL, plotting with matplotlib, manipulating data (numpy and pandas), and making reproducible reports in Jupyter notebooks. Data Science for Crime Analysis is also [available in Spanish and French](https://crimede-coder.com/blogposts/2025/BookTranslations). Both of the books are highly relevant even for people outside of crime analysis. The LLMs for Mortals is relevant for every software engineer, and is the only book on the market with major examples across all model providers that is up to date. Data Science for Crime Analysis has crime analysis examples in the book, but is really a better introduction to environment and data management than other books currently on the market. ## Services The consulting services I offer are: ### Program Analysis Is your program causing crime to go up or down? CRIME De-Coder uses state-of-the-art causal inference techniques to know how well your program is working. See ### Predictive Analytics Predictive analytics, aka machine learning or artificial intelligence, is using statistical models to predict the probability of a future event occurring, or forecast a numeric value that is likely to occur in the future. Examples are predicting the number of crimes likely to occur in an area in the future or predicting the probability an individual is going to be involved in gun violence in the next year. These are not like predictions in Minority Report, they do not call for preemptively arresting people before they have committed a crime, but uses different proactive tactics to prevent crime before it occurs. CRIME De-Coder has extensive experience in predictive analytics, has published several papers and won multiple prediction competitions (NIJ recidivism forecasting, NASA Algae Bloom). Examples include Spatial Based Forecasts, Person Based Predictions, Fair Applications of Predictive Analytics (fair algorithms), and Time Series forecasting and monitoring. ### Process Automation Examples of process automation to help police departments and other agencies include: - Data processing (extract, transform, load data into different databases, like ESRI tools) - creating dashboards (Tableau, PowerBI, custom D3.js, ESRI) - translating old code to open source (e.g. SPSS or SAS to python or R) - custom software (creating a master name index, police contact surveys, mapping applications) ### Civil Litigation I provide data analysis for civil litigation cases, these include: - crime analysis related to premises liability - racial bias analyses - fourth amendment rights violations with police technology I have experience in cases across the country, have been deposed and considered an expert multiple times, and provide consulting services to both defendants and plaintiffs across each of the areas. You get an independent expert assessment when you contract my services. ### Crime Analysis Training I was a former professor of criminology, and have conducted trainings across the country. These include short conference trainings, multiple day in person trainings, and traditional college courses over a semester. Example topics include: - python training (two day training delivered to Austin and Kansas City) - time series analysis (custom course developed for online learning platforms, and at the IACA conference) - Geographic Information Systems (college course using ESRI tools taught at SUNY Albany and University of Texas at Dallas) If you want more technical training for your analysts, Andrew Wheeler and Crime De-Coder is your best option. ### Workload Analysis See . Workload analysis for police departments can span different areas, such as: - Do you have the right number of officers to effectively respond to calls for service? - Are you spatially allocating officers to minimize response times? - Are you effectively triaging detective case assignments? These questions often mix advanced statistical analysis, discrete event simulations, and optimization models (e.g. linear programming), to develop strategies, policies, or empirically justified forecasts to better meet the demands for your police department. Redistricting: Response times are critically important for public safety – faster response times by police and emergency services can make the difference between life and death (Circo & Wheeler, 2021). The CRIME De-Coder has written custom applications to reduce drive responses to calls for service, as well as to rebalance workloads via redistricting (Wheeler, 2018). Implementing Gun Violence Reduction Initiatives: The CRIME De-Coder has worked with multiple police departments implementing gun violence reduction initiatives. These include programs like Focused Deterrence, or chronic offender initiatives. CRIME De-Coder has developed custom social network analysis methods (Wheeler et al., 2019), and has experience in linear programming to identify optimal allocation models. ## Demos I have several demo's on my site to demonstrate my technical expertise in generating software, see them at [Crime Trends Dashboard](https://crimede-coder.com/graphs/Dallas_Dashboard): Fully automated dashboard with up to date Dallas crime trends. Tech stack: serverless dashboard using pyscript and automated daily refresh using github actions. [WDD Crime Reduction Tool](https://crimede-coder.com/graphs/WDD): A html and javascript application to apply a statistical test CRIME De-Coder developed to estimate how many crimes were reduced in a place based intervention. [Interactive Hot Spot Map](https://crimede-coder.com/graphs/DurhamHotspots): Customized, interactive mapping applications. Tech stack: leaflet and python/folium to create customized looks and interactions. [Network Prioritization Tool](https://crimede-coder.com/graphs/network): Upload your own network and apply algorithm to prioritize gang call-ins. Tech stack: interactive javascript application using sigma.js. [Custom Survey](https://crimede-coder.com/graphs/survey?surv=se1): Demonstration of how CRIME De-Coder can implement customized surveys. Tech stack: PHP with google sheets backend. [Sworn Dashboard](https://crimede-coder.com/graphs/sworn): Dashboard to compare policing hiring statistics to other agencies in the US (based on data from the FBI). Tech stack: Supabase and SQL functions to generate statistics, and D3.js. [Example Static Report](https://crimede-coder.com/images/report.pdf): Sometimes all you want is a static report. This auto-report is created and refreshed once a week using python, Quarto, and word (and hosted on web as PDF). [GSU Book Prices Table](https://crimede-coder.com/graphs/GSUBooksQuery): Dashboard showing how to conduct ad-hoc queries. Uses Cloudflare R2 to host the data, DuckDB WASM to form ad-hoc SQL queries, and D3.js to create a sortable table. ## Blog My blog home page is , and I have an RSS feed of new blog posts at . I often write about somewhat technical topics related to crime analysis, but often in an accessible way. For example blog posts: - [Coding Agents Risk Leaking Secrets](https://crimede-coder.com/blogposts/2025/CodingAgentRisks) (Saying how agents can leak credentials, and potential mitigations like IP scoping secrets) - [Crime Analysts make too many BOLOs](https://crimede-coder.com/blogposts/2025/BOLOs) (how sending PDFs of BOLOs are often not read, and discuss better alternatives for crime analysts time) - [How to use python to query your RMS](https://crimede-coder.com/blogposts/2024/QueryRMS) (a tutorial on running SQL inside of python) - [Million Dollar Hotspots](https://crimede-coder.com/blogposts/2023/MillionDollarHotSpots) I have many more (approximately one per month). ## Newsletter I have a newsletter on helping students pursue private sector careers, https://crimede-coder.com/newsletter, various tech advice (secrets in python, SQL CTE's, basics of LLMs, how to run code from the correct directory), how to pursue careers (banking, healthcare, specific companies), and examples of individuals with PhDs in social science that are now in the private sector and examples of what they work on. ## Presentations I post presentations to . If you want me to present to your group, get in touch! Some example prior presentations are: - [Avoid Chasing the Noise](https://crimede-coder.com/present/Penn_AvoidNoise_Wheeler.pdf) (talk at UPenn) - [Simple Stats for Evaluating Crime Reductions](https://crimede-coder.com/present/VCAN_SimpleStats_Wheeler.pdf), slides for Webinar for Virginia Crime Analysis Network - [Tech Talk on License Plate Readers to Libertarian Party of Wake County](https://crimede-coder.com/present/CrimeDecoder_LPRs.pdf) - [Mapping the Risk Terrain for Crime Using Machine Learning](https://crimede-coder.com/present/Wheeler_FBI_2023.pdf) FBI Dash conference I am interested in doing presentations for my new book, Large Language Models for Mortals, so reach out if that is something your group would find useful. ## Other Site Information You can see the entire set of pages that are publicly available at or . You can follow my work at LinkedIn is the best to keep up on new material,