Deeplocal is not an ad agency, not a product company, and not quite a software firm. The Pittsburgh-based studio invents things: interactive campaigns, physical robots, connected experiences, and custom hardware projects that brands like Google, Nike, and Hershey have paid serious money to put in front of the world. It builds out of a converted space in Lawrenceville, and it has done so since 2006, when founder Nathan Martin left Carnegie Mellon's Entertainment Technology Center to start something he couldn't quite name yet.
That's the short version. The longer one is more interesting.
What Deeplocal actually does
The studio sits in a category it more or less invented: creative technology. Projects range from a Google Auto-Draw installation that let strangers sketch together across city blocks to a Nike shoe customizer that used real-time athletic data to generate unique colorways. Deeplocal built a Nutella-dispensing brunch robot. It created a live, crowd-powered basketball shot for ESPN. Each project starts as a brief from a brand, gets filtered through a team of engineers, designers, fabricators, and writers, and comes out the other side as something you can touch or interact with.
The studio doesn't run ad campaigns or manage media buys. It builds the thing itself, handles engineering and fabrication in-house, and ships it to wherever in the world it needs to live. That full-stack capability is the competitive edge. Most agencies have to outsource the physical build. Deeplocal doesn't.
The Carnegie Mellon connection
It's hard to overstate how much of Deeplocal's DNA traces back to Carnegie Mellon. Nathan Martin came out of the Carnegie Mellon ETC, the graduate program in interactive media that has quietly shaped a generation of creative technologists. That program's blend of storytelling, engineering, and rapid prototyping maps almost exactly onto how Deeplocal approaches a project brief.
Several of the studio's early hires came from CMU, and the university's broader Oakland ecosystem, with its robotics labs and human-computer interaction programs, gave Deeplocal a local talent pool that would be difficult to replicate outside Pittsburgh. The city's tech corridor has been shaped by a few quiet giants, and that CMU pipeline runs through most of them. Aptiv's engineering presence in Pittsburgh draws from the same reservoir.
Why Pittsburgh, why Lawrenceville
Martin has been asked the Pittsburgh question a lot. The answer isn't romantic. It's practical. Real estate costs in Lawrenceville allowed Deeplocal to lease the kind of high-ceiling fabrication space that would cost three times as much in Brooklyn or San Francisco. The studio needs room for CNC machines, laser cutters, and prototype hardware. A 400-square-foot open office does not work.
The neighborhood helped, too. By the time Deeplocal was growing its team, Lawrenceville had enough density of designers, makers, and curious technical people to make recruiting feel less like a long shot. The murals, the studios, the workshop culture: all of it contributed to a place where a team building weird robots for Nike didn't feel out of context.
Projects that put the studio on the map
Three projects in particular pulled Deeplocal out of regional obscurity and into international press coverage.
The first was Google's "Auto Draw" activation, which used machine learning to interpret rough sketches in real time. Deeplocal built the physical installation layer, turning a software product into a public experience people lined up to try. The second was the ESPN SportsCenter challenge, where the studio wired a basketball hoop to a live broadcast feed and let viewers at home control a robotic arm taking the shot. It worked, live, on national television. The third was a Hershey project that involved a custom-built machine dispensing precise chocolate servings based on individual visitor data inputs. All three required the same capability: engineering, software, fabrication, and a concept strong enough to earn press on its own without a paid media campaign behind it.
How the studio is structured
Deeplocal runs on a studio model rather than a traditional agency structure. Project teams form around a brief and include engineers, industrial designers, software developers, and a project strategist. There is no dedicated account management layer separate from the creative team. The people pitching the idea are the same people building it.
That model keeps client relationships honest and the project scope tight. It also means the studio tends to take on fewer projects than a comparably sized agency, because each one demands genuine engineering hours, not just concepts and decks.
What the studio means for Pittsburgh's tech identity
Pittsburgh's tech conversation tends to cluster around autonomous vehicles, robotics research, and university spinouts. Deeplocal fits none of those categories neatly, which is part of why it gets less attention than it deserves in the city's own coverage of itself. It's a fully commercial creative business that exports Pittsburgh talent and Pittsburgh craft to global clients, generating revenue, jobs, and reputation without relying on grant funding or academic partnerships.
That matters for the city's self-image. Pittsburgh produces more than research. Deeplocal is evidence that a small studio, started by someone who stayed here instead of leaving, can compete for the same briefs as firms in New York and London. It's been doing exactly that for close to two decades, from a building in Lawrenceville, with a team that mostly grew up nearby.
That's not an accident. It's a model.