Work
Customer data, treated like infrastructure.
I turn customer data into something a company can actually use. Most companies collect more data than they can trust. I close that gap where it actually breaks: how data gets collected, how its quality gets enforced, and how it gets activated across marketing, personalization, product, and AI.
I’m an early adopter and a skeptic at the same time. I put new technology into production, and I stay skeptical enough to ask where it actually earns its keep. That’s the edge in a tech sector mired in one-sided thinking: hype on one side, dismissal on the other, and not much judgment in between.
Data quality. I build the layers that make data trustworthy: identity resolution that gives you one profile per person instead of one per system, governance that keeps collection clean at the source, and automated checks that catch breakage before anyone makes a decision on bad data. AI is only as good as what’s underneath it. I make sure what’s underneath it holds.
Collection. I design how customer data enters the building: one governed way to record events, identity attached automatically, personal data kept separate. When collection is an afterthought, every downstream team pays for it. I make it a standard, so personalization, ML, and analytics all start from truth instead of janitorial work.
Activation. Data nobody uses is shelfware. I build the audience systems and integrations that turn a customer data platform into something the whole company actually reaches for: self-serve, documented, and adopted because it’s the easiest path, not because someone mandated it.
Patterns others miss. Part of the job is seeing what the dashboards don’t say: the slow drift in data quality, the integration everyone assumed was working, the number that moved for the wrong reason. I go looking for the thing nobody’s checking.
Grounded decisions. I decide on evidence, not on who’s loudest or what’s fashionable. Vendor evaluations run across legal, marketing, data science, and product, and they land because the criteria were honest. I set multi-year direction the same way: write down the reasoning, so the decision survives the meeting it was made in.
Process and culture, early. Systems break, but process and culture break first and quieter. I catch it early: the review bottleneck that’s really a trust problem, the tribal knowledge that’s really a documentation problem, the team waiting on one person for everything. Then I act on it. Automated checks instead of gatekeepers, standards instead of folklore, and people developed for autonomy rather than dependence.
The work spans technology, people, and culture, because in practice they’re the same job. Recent work includes a vendor-agnostic identity layer behind a Marketo-to-Iterable migration and an org-wide Tealium-to-Hightouch re-architecture, cut over product by product without disruption. The goal is the same every time: data the organization can trust, systems that run without me, and people ready to own what’s next.
← All chapters
