Hi, I'm Michael
I lead data operations at the Child Poverty Action Lab, a backbone organization for nonprofits across North Texas. CPAL works on the things that actually change child poverty in Dallas: public safety, maternal health, benefits delivery, housing, and criminal justice. I run our data engineering team, the data platform (Databricks, AI-enabled workflows), and the internal tools that put data in front of non-technical staff and partner orgs.
It took a few detours to get here. My first real job was coordinating a child development lab at UT Dallas, and the part that stuck with me wasn't the research itself, it was sitting with the parents of our young participants and trying to explain what we'd found. The gap between how researchers talk to each other and how findings actually reach the people who can do something with them turned out to be the thread of my career. It pulled me through a policy internship at Children at Risk, an MPP at UT Dallas, and into CPAL in 2020.
Two convictions sit underneath the work. First: public data is only useful when the people closest to a problem can act on it, and most of the data that could help families is still locked up in formats and systems that field-team workers, community organizations, and decision-makers can't reach. Second: nonprofits deserve the same data infrastructure as the institutions and markets they're often working against. Most of what I do, at CPAL and outside of it, is in service of those two ideas.
Public data is only useful when the people closest to a problem can act on it.
What we've built at CPAL
A snapshot of the work, from the decisions it informs to the infrastructure underneath.
- 01 A 30+ Shiny app suite informing decisions across CPAL focus areas: housing, public safety, maternal health, benefits delivery, and criminal justice.
- 02 An eviction data workstream feeding newly-filed cases daily to 12+ partner organizations across legal aid, advocacy, government, and academic research; the data behind North Texas Evictions.
- 03 Internal tools that let non-data staff act on data without analyst intervention, including a parcel-level outreach tool with 20 active field-team users.
- 04 The Databricks platform migration on AWS, moving 35-40 pipelines from file-based storage onto unified cloud infrastructure with Git-tracked orchestration.
- 05 AI-assisted workflows I've built into how I develop pipelines, write documentation, and review code (Claude Code with custom skills and agents), now rolling out across the team.
- 06 A data engineering team scaled from 1 in-house analyst to 6 external data engineers + 1 in-house data engineer, executing on the internal data roadmap.
Selected Media & Visualizations
Featured in
- D Magazine The Lawyer Who Landlords Don't Want to See in Court May 2024
- KERA News Eviction less likely for Dallas County tenants who get a lawyer, but most don't have one January 2024
Data visualizations contributed to
- The Lab Report How Dallas Police Ramped Up Homeless Enforcement April 2026
- The Lab Report Where Did the Patrol Cops Go? February 2026
Technical Skills
Analysis & Languages
- R (tidyverse, sf, Shiny)
- SQL
- Python
- TypeScript
Visualization & Geospatial
- R Shiny
- Tableau
- Mapbox GL
- Spatial SQL
- QGIS
- ArcGIS
Methods
- Risk Terrain Modeling
- Index Construction & PCA
- Spatial Analysis
- Needs Assessment & Supply Modeling
Domains
- Housing & Eviction
- Public Safety
- Criminal Justice
- Maternal Health
- Benefits Delivery
- Community Development
Data Platform
- Databricks (Lakehouse + Unity Catalog)
- PostgreSQL / Neon
- DuckDB
- Polars
- PostGIS
- CKAN
AI Workflows
- Claude Code (custom skills & agents)
- Anthropic API
- Google Cloud Vision (OCR)
Orchestration & Infra
- Terraform
- GitHub Actions
- Docker
- AWS