In the previous article in this series, we examined Mawari’s role in bringing AI Agents to life. But an important question flowing from that narrative is how? This was covered to a degree, but now it’s time to go deeper. 

To do so requires asking a few questions. What is Mawari? Who stands behind it? How did it get here? And most importantly, how is it positioned – in timing and technical competency – to achieve Digital Humans and several other endpoints in its ambitious road map? 

The TLDR version is that Mawari is positioned as a key XR enabler. Though the Metas and Apples of the world get more credit, behind-the-scenes players like Mawari unlock XR’s real potential. As tech pundits say, selling picks and shovels in the gold rush is an opportune role.

Mawari is well on its way with a seven-year tenure and recent strategic fundraising. It also has the rare distinction as a revenue-generating player in the still-nascent spatial computing realm, as they generated $1.5M+ in revenue during 2024

As for the who? Its team collectively boasts 3 major exits, 50+ years in computer graphics, AR, and networking. It also has ample pedigree, considering talent from Meta, Google, Magic Leap, Foundry, Grayscale and FalconX. This cross-disciplinary firepower is aligned with its road map. 

But how did it get to this point? Let’s pull on that thread. 

Balancing Realistic and Imaginative

Before examining Mawari’s backstory, a bit of context is required to truly understand the value that tech enablers and infrastructure players like Mawari bring to the table. As an analyst, one of the mistakes I’ve seen made repeatedly in tech cycles is too much imaginative future gazing. 

Sure, it’s a good thought exercise to imagine emerging-tech use cases. But what’s often missing is acknowledgment of the technical realities that stand in the way. This does a disservice because it overpromises and, in turn, erodes credibility and confidence in the technology. 

So how do we avoid such outcomes? One is to acknowledge technical challenges while discussing XR’s imaginative endpoints and to refrain from overpromising. The other is to actually tackle those technical challenges and remove them as impediments to XR’s growth.

That second action is more impactful and is currently being executed by tech giants like Meta and Apple. As a future-proofing move to establish platform positions, they’re investing tens of billions annually to accelerate the process and get more devices into people’s hands & heads,

But even if we get the hardware and holography dialed in, users adopting en mass, rich content libraries, and all the other requirements listed above, another challenge awaits: reliable content/experience delivery and stable infrastructure. That brings us back to Mawari. 

Missing Piece

Mawari’s superhero backstory begins in 2017 with a vision from founder Luis Oscar Ramirez to be a technological enabler for the XR revolution that everyone was talking about – including much of the “future-gazing” rhetoric noted above. He saw a critical missing piece in these visions. 

That missing piece was a way to reliably deliver all the content and interactions for rich XR experiences. We’re talking about bandwidth-heavy experiences that won’t perform to their expected levels if delivered over legacy infrastructure that was built for 2D content.

Just like XR hardware and software are purpose-built, their infrastructure needs to be likewise native and specialized for their unique and extensive needs. Those needs include low-latency and performance, such as dimensional interaction between digital and physical subjects. 

Meanwhile, streaming services like Netflix have set a standard by conditioning the world to expect reliable content delivery. As we continue to evolve from 2D to 3D, XR will inherit those expectations. So there isn’t room for error, nor consumer tolerance for subpar streaming quality. 

This all translates to the need to eliminate bottlenecks and boost throughput. This traditionally means widening the pipe and/or narrowing the content. The former involves things like edge computing, while the latter is historically addressed by compression technologies such as mp3.

The MP3 of XR

So Mawari set out to solve these problems – and several orbiting issues – starting with compression. A key question that Ramirez faced in early days was what will be the mp3 of XR? This led to Mawari’s foundational innovation: a codec for lossless 3D file compression. 

Then, the second leg of the stool was a spatial rendering and streaming pipeline. Packaged into Mawari’s Spatial Streaming SDK, it renders 3D content in the cloud and delivers it efficiently to devices. It does so by assigning the burden of processing and rendering to the cloud. 

This approach taps into the beefier memory and computing resources of cloud servers versus underpowered smartphones. Experiences are then streamed to devices – utilizing the compression codec noted above – resulting in reliable and load-balanced XR experiences.

For example, 3D elements stream from the cloud while devices can handle other tasks such as scene mapping that ensures XR experiences interact realistically and dimensionally with physical spaces. Altogether, this split rendering balances the computing load optimally. 

As an SDK, these capabilities also achieve scale by being packaged up and delivered to XR developers to empower their delivery and distribution capabilities. But to truly unlock Mawari’s potential, it needed another key component in its product mix: a network.

Picks, Shovels & DePIN

That brings us to the present and to the final piece of the puzzle: The Mawari Network. As we teased in the previous part of this series, it relies on an emerging construct: DePIN. Short for, Decentralized physical infrastructure networks, it could be the key to unlocking XR’s promise.

The Mawari Network applies this principle through decentralized and geo-diverse GPU-powered nodes. Mawari’s Guardian Node license enlists unused compute resources from individuals who collectively power the network, while incentivized through token-based compensation.

This approach avoids a bottleneck in traditional networking. In these legacy systems of centralized servers, there aren’t enough GPUs at the network edge to accommodate XR’s throughput requirements. This results in a lack of geographic proximity to stream XR globally. 

That last part is critical because XR’s heavy throughput means that it can only be streamed if within a certain distance to the closest GPU at the network edge. And that coverage doesn’t exist today. Even AWS and Google Cloud don’t have the GPU capacity for global coverage.

And they don’t have the incentive to build it, given relatively-low XR adoption. So this lack of GPU coverage at the network edge is something that Gonzalez identified as a gating factor to XR’s promise; and to Mawari’s ability to deliver its Spatial SDK. So he decided to build it. 

Spatial CDN

When you add it all up, the Mawari Network is a sort of decentralized content delivery network (CDN). This is an established networking principle on the 2D web, but modified for spatial computing. It represents the purpose-built and native requirements examined earlier. 

And it’s already working. Mawari reports that its decentralized network reduces bandwidth by about 80 percent. That in turn has a direct and tangible impact on cost, performance, and reliability. Its demand, and therefore value, will only grow as spatial computing itself does. 

But the remaining question traces back to a word we keep using as a reality check in this series: how? To propel the Mawari network, it will host a node sale. Coming up very soon, users globally will be able to join the network, strengthen its capacity, and make money doing it.