Patrick Noonan

I build the data models, pipelines, and dashboards teams rely on.

Analytics engineer with 15+ years in data. I work mostly in SQL, Python, dbt, Looker, Omni, Databricks, and Redshift, and I get up to speed quickly on whatever BI tool or warehouse your team already uses. Available for contract and fractional work.

U.S. citizen based in Spain. Working hours overlap 6 AM to 3 PM Eastern.

Email me View LinkedIn
Diagram of data flowing from raw sources through staging and intermediate models to fact and dimension models, then to dashboards Raw sources Staging Intermediate Facts and dimensions Dashboards

What I take on

  • Building, owning, or cleaning up a dbt project

    Day-to-day model development and ownership, plus the conventions, testing, and governance that let the team trust and extend it.

  • Dashboards built around the question

    I start with your team or your clients: what decision they’re making and what they need to see. Then I build the metrics and dashboards that answer it.

  • Looker and Omni semantic layers and migrations

    LookML and Omni modeling, administration, and moving from one BI tool to another without losing what people rely on.

  • Pipeline architecture on Databricks and Redshift

    Ingestion, transformation, automated syncs, and alerting when something fails.

  • Moving off dbt Cloud to dbt Core

    Scheduling, CI, and deployment on infrastructure you already pay for.

Selected work

Client and employer names are left out on purpose. I’m happy to talk through any of these in more detail.

Compliance reporting pipeline for a state health department

A state health department client needed newborn-screening data modeled to CDC EHDI standards, and its compliance reporting was done by hand.

What I built
A Databricks pipeline and dbt models for the screening data, using Delta Sharing, with automated sync and same-day failure alerting. On top of it, three client-facing Omni dashboards.
Result
The dashboards replaced the manual compliance reporting and became the client’s primary reporting tool.
Tools
Databricks, Delta Sharing, dbt, Omni

Looker to Omni migration

A healthcare software platform whose dashboards served health workers at 800+ physical sites. The goal was to cut BI costs without disrupting the clients who relied on those dashboards.

Choosing the tool
Evaluated three to four BI tools and chose Omni for its features, functionality, and semantic layer.
Scoping
Pulled reports from Looker to map the existing reporting, then worked with product managers and client services to decide what to port and what to retire.
Building
Rebuilt the semantic layer in Omni from the LookML, moved logic out of Looker PDTs and into dbt, and recreated the client dashboards.
Rollout
Led client trainings so users moved over smoothly.
Tools
Looker, LookML, Omni, dbt

A dbt project built from the ground up

A dbt ecosystem for the analytics team at a healthcare software platform.

What I built
The team’s dbt project from scratch, with modeling governance and testing practices. It now covers 290+ models and 80+ tests.
Result
Faster dashboards and queries, from implementing Looker persistent derived tables (PDTs) and re-architecting the LookML models.
Tools
dbt, LookML, Looker PDTs, Redshift, Databricks

dbt Cloud to dbt Core migration In progress

The dbt Cloud Enterprise plan cost five figures a year, and the team already paid for the infrastructure that could replace each piece of it: Databricks, GitHub, and AWS.

Scheduling
Databricks Workflows’ native dbt task replaces the dbt Cloud scheduler, with daily, weekly, and monthly runs split by tag.
CI
GitHub Actions builds and tests only the models each pull request changes, in a temporary schema, compared against a production manifest kept current in S3.
Validation
Confirmed parity between dbt Cloud and dbt Core with automated tests and by comparing the manifest before and after the switch.
Result so far
CI has already caught a real bug before merge: a column removed from one dimension table silently broke a model three hops downstream.

Writing

About

I’m an analytics engineer based in Dénia, Spain, with 15+ years in data. I started at Pandora Internet Radio, growing from ad operations into a senior data analyst role supporting FP&A. Since 2020 I’ve worked fully remote, both on larger data teams and as a team of one, owning the stack end to end. That means wearing a lot of hats: data engineer, analytics engineer, analyst, dashboard builder, and client-facing partner.

The work I enjoy most is bringing structure to messy data: defining metrics and the governance around them, automating the manual processes teams get stuck with, and setting up data quality alerts so problems get caught before they reach a dashboard. With stakeholders, I like building what actually helps: sometimes a full dashboard that informs their decisions, sometimes just a scheduled CSV export that replaces a manual step.

I’m a U.S. citizen, my hours overlap with U.S. Eastern from 6 AM to 3 PM, and I’m comfortable with 1099 contract arrangements. For nearly five years I also mentored data analytics students at CareerFoundry, and I helped dozens of them land data jobs.

Portrait of Patrick Noonan

Contact

If your team needs help with a dbt project, a BI build or migration, or a data pipeline, send me a note with what you’re working on.

Email me View LinkedIn