Fractional Data Engineering

Messy, fragmented data — turned into pipelines you can trust.

I'm Will Jewell, a senior data engineer. I give small teams a real data-engineering function — clean multi-source pipelines, streaming ingestion, and reporting people actually rely on — without the cost or wait of a full-time hire.

Book a 30-min intro See how I work
Python + SQL · ETL Kafka + HL7 · streaming multi-EHR · normalization Prometheus / Grafana · monitoring Tableau + Power BI · reporting
12+ yrsowning the full data-delivery lifecycle
180+hospitals served by pipelines I built & ran
~20Kdaily users on the clinical app they feed
76→97%data quality lifted on a multi-million-row audit

For a lot of teams, the data itself is the bottleneck.

The numbers live in five systems that don't agree. The pipeline breaks quietly and nobody notices until a report is wrong. Hiring a full-time data engineer is slow and expensive, and the work needed doesn't fill a permanent seat yet.

That's the gap I fill. I plug in as a fractional data engineer — senior, hands-on, and fast — and stand up the pipelines, integrations, and reporting your team needs now. You get the data-engineering function without the overhead of a full-time hire, and you keep everything I build.

What I do

Senior data engineering, delivered fractionally.

The messier and more fragmented the data, the more useful I am. Common engagements:

Multi-source ETL & pipelines

SQL databases, APIs, files, and ERPs pulled into one clean, well-modeled dataset — scheduled, tested, and monitored so it doesn't break in the dark.

Healthcare data integration

Normalizing data across multiple EHRs and claims sources into consistent, analysis-ready models — the kind of messy, high-stakes integration I've done at scale.

Streaming ingestion

Real-time and event-driven pipelines on Kafka, so data lands where it needs to be as it happens — not on tomorrow's batch.

Warehousing & modeling

Snowflake and SQL warehouses set up and modeled properly — so analytics are fast, trustworthy, and cheap to query.

BI & reporting

Dashboards and reporting your stakeholders actually rely on — built on a foundation that's correct underneath, not just pretty on top.

AI-accelerated delivery

I build with AI tooling in the loop, which means a fractional engagement ships far more than the hours suggest — without cutting corners on rigor.

How I work

Production is the finish line — not a slide deck.

A tight, transparent loop. You see working software early and own it at the end.

01

Frame

Understand the data, the systems it lives in, and the decisions it's supposed to feed. Scope the smallest thing that's genuinely useful.

02

Build

Pipelines as versioned, tested code — not fragile scripts. Reviewable, reproducible, and built to be handed off.

03

Ship

To production, monitored, with alerting when something drifts. The value shows up in your systems, not a report about them.

04

Hand off

Documented and explained, so your team can run and extend it. No black boxes, no lock-in to me.

PHOTO
your headshot
drops in here
About

I've spent 12+ years turning fragmented, high-stakes data into things people can trust.

I'm currently a Senior Data Science Engineer at HCA Healthcare, where I design and run the data pipelines behind NATE — a clinical decision-support app used by ~20,000 people a day across 180+ hospitals — pulling from a complex, multi-EHR environment and holding it to real SLAs with Prometheus/Grafana monitoring and source-to-output lineage.

Along the way I built NATE's streaming foundation — a microservice that parses HL7 clinical messages off a Kafka feed into a standardized model — and earlier ran a top-10 Tableau site at HCA with 40+ dashboards, including a Meaningful-Use compliance dashboard tied to $220M+ in federal incentives. Before healthcare I built SQL models for 7 refineries across 3 countries and audited a multi-million-row database from 76% to 97% data quality. I trained at Georgia Tech (computational data analytics; B.S. engineering, President's Scholar).

The through-line is the unglamorous, essential part: getting messy inputs into a clean, reliable state so the analytics on top are actually dependable. I take on select fractional engagements because a lot of great teams need that function long before they need a full-time seat.

What I care about: outside work I'm a dad, a Raspberry-Pi tinkerer, and I stock a community fridge. I especially like working with teams in climate, health, and civic tech — mission-driven work where getting the data right actually moves the needle.
Engagements

Two simple ways to work together.

Fractional retainer

An ongoing data-eng function

A set slice of my week, every week. Best when the data work is continuous — new pipelines, integrations, fixes, and reporting as they come up.

  • Predictable monthly cost, no full-time overhead
  • Senior hands from day one
  • Scale up or down as needs change
Project / SOW

A defined build, shipped

A specific outcome with a clear scope — a warehouse migration, a new ingestion pipeline, a reporting layer. Fixed and delivered to production.

  • Clear scope and deliverable up front
  • Shipped, monitored, documented
  • Your team owns it at handoff
Let's talk

Got data that's more headache than asset?

Tell me what you're wrestling with. A 30-minute call is usually enough for me to point at what I'd do first — whether or not we end up working together.

Book a 30-min intro Email me