Work With Me

Available for project work and the right long-term role, in northwest Georgia or remote.

1. Measure first. Know the baseline before changing anything. It’s where the company name comes from.
2. Build it to last. Tested, logged and under version control. When something breaks, it says so before your users do.
3. You own it. Everything is documented, so your staff or the next consultant can carry it forward. I don’t hold customers hostage.

What I do

#

Build the database behind your reporting

#

Good reporting sits on a well-designed database. I design the tables, views, stored procedures and functions that turn raw application data into something people can query with confidence. T-SQL is my strongest suit.

• Data marts and reporting layers over EHR, ERP and line-of-business systems
• Set-based rewrites of slow procedures; window functions, recursive hierarchies
• A standard header on every object: what it does, how to call it, what changed and when

Clean and transform data between systems

#

Your data lives in a cloud application, an on-premises database and a partner’s nightly file. Somebody has to corral it into one place, on schedule, without anyone retyping it. In cowboy-technical terms, you need a data wrangler.

• Scheduled ETL with SSIS and T-SQL, including SFTP pickups and partner extracts
• One-time migrations when you change vendors
• Cleanup: de-duplication, address standardization, reference data
• HL7 interfaces with Mirth Connect for healthcare

The payoff is usually staff time. Work that was tedious and error-prone runs on its own, and the data comes out more accurate. For an example, see Census Exports.

Turn your data into dashboards people use

#

Dashboards get used when a number means the same thing to everyone who reads it. I start with agreed-on definitions, kept in one place as a Power BI semantic model with proper relationships and DAX measures. Self-service is then safe, instead of every person for themselves. And there is still nothing like the “Aha” moment when someone sees their own data clearly for the first time.

• Power BI semantic models and DAX
• Operational and clinical dashboards: capacity, readmissions, revenue cycle, COVID-19
• Paginated reports for when it has to print

How I work

#

Documentation is the step most often skipped on data projects. For me it’s a matter of business continuity and trust: anything I build, a knowledgeable person should be able to understand and extend without me. That has probably cost me some revenue over the years. I’m fine with that.

In practice:

• Version control for database schemas and SSIS packages, in Git
• Logging inside batch jobs, so a failed step points to itself
• Unit tests for data that run daily and email the right person when something looks wrong
• Metadata harvesting that documents what’s in the databases and Power BI reports, automatically
• A wiki written for the next person, with pointers to it from the code

Track record

#

• More than thirty years designing Microsoft SQL Server databases
• Twenty-four years running my own firm, including an electronic medical records system used by 60 agencies in 6 states
• Senior BI developer for a regional health system: led its move to Azure DevOps and Git, and wrote most of the 640+ objects in its main reporting data mart
• Customers in healthcare, manufacturing, public schools and service businesses

The Projects and Guides show the work itself. The About page has my background.

Questions people ask

#

Get in touch

#

Tell me a little about what you’re working on. Your details come to me and nowhere else: no mailing list, no sales sequence.

Prefer LinkedIn? Connect with me there.