A founder-led data firm.
You hire us, you get us. Senior talent, always hands-on. No account managers, no offshore handoffs, no junior engineers learning on your dime.
Untitled, on purpose.
Untitledis what it said on the new directory folder.
The problem is we, actually, care very much about names and calling something the right thing because it then drives coherence every time you use it. Or, it drives incoherence and is forever a thorn in your side.
We realized quickly that Untitled actually represents the way we give ourselves to the work. Every project arrives with its own shape. Untitled is how we show up to it: full attention, no assumptions. Expertise without ego.
The Research bit came easier. We learn something new on every project and we take it with us. Decades of that.
Untitled Research is us. And we'll find the right name for the work we do for you.
We learn fast so you move faster.
We're a services firm grown out of a research lab, not a product company that does services on the side.
Between engagements we build our own things: internal tools, prototypes, product experiments. Some ship. Some don't. Either way, every experiment teaches us something valuable and real, and hones our skills.
The data landscape and tooling changes drastically every few years. We stay sharp by building, not by reading about it. We try, we fail fast when we should, and we keep going. Enthusiastic by default, open-minded by design: it's the only way to stay current in a field that never stops reinventing itself.
What we keep coming back to.
On the industry
There is no magic.
There's just good work, which is categorically different from bad work. The industry sells magic: AI that transforms, platforms that supercharge, dashboards that delight. But none of it works without people who understand the problem. Tools amplify, they don't substitute. Good work is patient, specific, and more concerned with details than taglines.
Watch the incentives.
Every platform, vendor, and tool you adopt has a roadmap that isn't yours. Every. One. Pricing tilts toward their margin. Egress fees compound. Proprietary formats outlast the contract that introduced them; open-source projects pivot, get acquired, or change their license. Adopt with eyes open and price the exit before you commit to the entry.
Technology is not the goal.
A model is meaningful only if it shapes a decision. A dashboard is impactful only if someone reads it. Real-time is just a massive cost if it can't affect real-time change. Technology is a means, never the end. You cannot buy your way out of data trouble and bad process. The same architecture has produced transformative outcomes in one organization and rot in another. What surrounds the tech matters more than the tech itself.
On the work
Names matter.
What you call your org unit, your metric, your project matters. It's a small speed bump to call a thing by two different names, but you hit this speed bump every time. With every name. With every person who has a different name for it. Fighting this dissonance while you're trying to build and need everyone on the same page is huge hidden friction. Naming discipline and consistency up front saves months of rework later.
A little organization goes a long way.
Data work is organizational work. How to organize concepts, entities, pipelines, namespaces — the strength of the higher-level concepts decides whether any of it stays comprehensible and enables effective incremental evolution. A well-organized foundation is a strong foundation. A disorganized one is a house of cards.
Building it is the easy part.
The first build grabs the headline, but the real work starts after the system goes live. Pipelines break, schemas drift, models go stale, the team that built it moves on. Most data systems fail not at launch but at month thirteen, when the original context has evaporated and nobody on the new team can tell what's load-bearing. Runbooks, docs, sensible defaults are all a downpayment on years of operation.
Two of us. Both senior. Both reachable.
Anthony
Data science and analytics, Cofounder
Data driven sherpa. Meets clients in their domain and language and develops a common foundation from which we can solution.
Goes to work when the alignment matters more than specs. Guides clients through the process, and makes sure we solve the right problem in a meaningful way, not just a technically interesting one.
Miro
Architecture and engineering, Cofounder
Literally a life-long entrepreneur. Self-starter; can solution his way around pretty much anything.
Built data applications when OS lived in a terminal and has not stopped building since. Owns the technical vision and architecture of every project we do.