DataBlock is one person.

I'm Edward Plata. DataBlock is my LLC, not an agency — there is no team behind this page, no account manager, and no junior who does the work after you've met the senior. You talk to me, and I do the building.

That is worth saying plainly because it changes what I'm good for. I take a small number of engagements at a time, which means I can go deep on a stack and genuinely learn a domain. It also means I can't staff a six-person delivery team, and if that's what a project needs, I'll tell you rather than stretch to fit.

What I work on

Databricks and Spark platforms for fintech, capital markets and Web3 teams. Medallion architecture, dbt governance, pipeline observability — the unglamorous layer that decides whether anyone believes the numbers coming out the other end.

The Web3 and Databricks overlap is a narrow one, and it's where I'm most useful: the data is public and adversarial at the same time, the tooling assumes a warehouse that doesn't exist yet, and the governance questions are the same ones a regulated fintech has to answer.

Between client work I build two products and run two autonomous systems that report their own numbers. Those are the honest portfolio — running software with costs and failure modes, rather than a slide about a methodology.

How I work

Handover is the deliverable. If your team can't run and change what I built after I leave, I've failed, however well it performed while I was there.

I'll say when something is a bad idea. Including when the thing you asked me to build isn't the thing that will fix your problem, and including when the honest answer costs me the engagement.

Small scope first. A two-week audit tells us both whether this is working, for a fraction of what finding out the slow way costs.

Open to forward-deployed roles

Most of this site is about contract engagements, but I'm also open to full-time forward-deployed and platform engineering roles at fintech, capital markets and Web3 companies — the kind where the job is to sit close to a customer's data, understand their domain properly, and ship against it.

That work suits me for the same reason the consulting does: the hard part is rarely the pipeline, it's knowing which number the business actually trusts, and why.

Want to talk?

If you have a platform that is expensive, untrusted, or about to be audited, that is the conversation I want.