The Track Record | AI for Construction Management

The path, in order.

Every line below is real work, told without the names.

2017

Marketing science.I started at a global media agency, on the analytics team embedded with one of the world's largest social platforms. Brand lift studies, ETL and reporting builds, and media optimization across programmatic, display, video, paid social, and search.

2019

First data hire.I joined a subscription nutrition company as its first data hire and built the warehouse, the reporting on top of it, and the forecasting models for inventory and new product demand as the company scaled. It was later acquired.

2021

Went independent.I founded a consulting practice in Los Angeles and have run it since: data infrastructure, eCommerce growth, and product leadership for consumer brands, with more than four years of business coming from referrals and repeat work.

Past two years

Multi-location growth.Third-party delivery and digital ordering for a restaurant group with more than 30 locations. I run the delivery channels, built the month-end close that grades ad spend market by market, and helped coordinate a dine-in digital ordering rollout across more than 50 people and 30 locations.

2026

AI, deployed.I moved from doing the analysis to building the systems that do it: document pipelines in production at a construction management firm, teams trained to run them, and live AI products of my own on the open web.

What the work looks like.

Three threads run through all of it.

Data infrastructure that outlasts me

I build the plumbing first. A SQL warehouse I built for two fast-growing consumer brands became the source of truth for strategic decisions at their global parent company. The same pattern repeats: automated pipelines in Python, and knowledge bases a team can query without me in the room.

A month-end close that ran by hand now runs from a documented pipeline anyone on the team can follow.

Growth across marketing domains

Media science, paid channels, eCommerce launches, and marketplaces. I have managed a team of three engineers, launched direct-to-consumer sales for a food brand, and served as tech lead for a retail brand's first store opening. For the restaurant group, third-party delivery sales grew 40 percent year over year in the best markets and 6 percent across all markets.

The ad framework behind that growth returned roughly $2.50 of incremental profit for every incremental ad dollar.

AI in production, not in demos

The construction work described on this site is running today at a firm in Los Angeles: a reporting pipeline in production, a bid comparison that leveled two bids sitting $1.3M apart on a $50M project, and eight generations of team training in under four months.

A 24-building campus inspection report went from raw PDF to a client-ready deliverable in under 10 minutes.

How the data actually gets used.

The data has never once told me what to do. Done well, it narrows a wide-open question down to a short list of decisions, each with a probability and a cost attached. A person still makes the call, owns it, and drives the strategy that follows.

Every result on this page happened that way. The warehouse gave a leadership team a shared set of facts to plan from. The ad framework showed which markets deserved money and which did not, and an operator made the cuts. The delivery growth came from people acting on clear numbers, month after month.

That is why the AI work on this site reads the way it does. The tools are faster now, and the shape of the job is the same: better inputs, clearer choices, and a person signing off on what goes out.