Jeff Ownby

Jeff Ownby

Your data stack, rebuilt and running on AI.

Available for contract work · Chicago · Currently shipping
5
AI Products Shipped
1.93M
Data Points Monitored
17
On-Chain Metrics
20+
Years in Data

Five focus areas. End-to-end. Built on 25+ years of doing the work.

01

Analytics Rebuild

GA4, GTM, server-side tagging, BigQuery as single source of truth.

02

AI Reporting

n8n + Claude pipelines. Reports write and deliver themselves.

03

AI Agents

Chatbots and AI characters with data-grounded knowledge bases.

04

Data Pipelines

BigQuery warehouses, Looker & Power BI dashboards, ingestion.

05

Full Products

When you need to ship a real AI product fast, end-to-end.

Three I built solo, end-to-end.

What the actual work looks like.

n8n workflow showing the monthly P&L pipeline
n8n · Claude API · BigQuery

The Monthly P&L pipeline.

Schedule trigger → pull BigQuery → write the executive narrative through Claude → assemble Word doc → upload to Drive. Runs every month with zero human touch.

BigQuery SQL view definition for sales projection
BigQuery · SQL

Inside one of the "Brain" queries.

Fiscal calendar logic, workday-prorated goals, channel attribution — the kind of view that drives the daily sales projection and FYTD goal-tracking system.

Server-side GTM Cloud Run dashboard
Server-Side GTM · Cloud Run

Server-side tagging in production.

sGTM running on Cloud Run with full observability. First-party FPID cookies, no client-side tracker for ad blockers to catch.

Looker Studio marketplace dashboard
Looker Studio · BigQuery

Marketplace performance dashboard.

ASIN-level fill rates, multi-account revenue trends, 12-month run rates against competitors. The kind of dashboard merchandising teams check every morning.

Winnie greeting players on Slotofun
Slotofun · AI character

Winnie on the daily spin.

Personality system writes contextual lines from session history and slot-industry trivia. (The Winnie on this site is the same character, moonlighting.)

AI Site Tour intake dialog with free-text and pre-built scenarios
Claude API · personalization

The AI Site Tour intake.

Built an AI-personalized site tour for an industrial client. Free-text input or pre-built scenarios ("We're seeing too many cut injuries on our glass-handling line"). The system tailors what comes next.

AI Site Tour personalized step with product recommendation and contextual messaging
Claude API · merchandising AI

One of the tour's personalized steps.

Custom messaging tied to the user's input. Specific SKUs, free-shipping math, ROI framing — generated for each visitor's actual situation. Real-time merchandising consultant.

Winnie commenting on a player's game session
Slotofun · in-session AI

Mid-game, reacting to real session data.

"Klondike's been decent to you" isn't a canned line — Winnie pulls actual session history and game state, then writes the comment.

AI is leverage, not delegation.
— the working philosophy

What you get with one builder + AI

  • Senior-team output at a fraction of the footprint. AI handles production; I run judgment.
  • Full-stack ownership — data architecture through AI surface. Not just one layer.
  • Working code, not slides. I configure, ship, test, automate, document.

When the laptop is closed

  • Watching soccer and the White Sox — with renewed hope, finally
  • Hiking, meditating, dancing at music events
  • Volunteering weekly at a food pantry
  • Traveling widely — has worked in Russia and Costa Rica

Ready to build it right?

Let's scope what fits — project-based, fractional, or fixed-scope.