If you’ve ever watched The Matrix Reloaded, Astro Boy, or Happy Feet Two, chances are you’ve seen the fingerprints of Yan Chen, though you’d never know it. A quiet force behind some of the most technically ambitious films in recent decades, Yan is the archetype of the “invisible innovator:” part computer scientist, part creative, all builder. Today, as Studio Architect for Kennedy Miller Mitchell (KMM) and CEO of AuroraAI, he’s redefining what VFX and animation pipelines look like in the era of real-time technology and generative AI.
Before that? He founded and exited LENS Immersive, an AI-powered video encoding company that raised $7.3M USD and developed streaming tools for platforms like Twitch and YouTube. And long before startups were cool, Yan was already setting up 3D animation studios for Disney and DreamWorks and supervising effects for Hollywood’s biggest names.
In this exclusive conversation, we sat down with Yan to explore his journey, his philosophies on emerging tech, and what’s next for content creation at the bleeding edge.
Dual passions, one mission
“I was torn,” Yan begins. “I studied neurology and computer science with an AI emphasis. But then part of me also just wanted to entertain people.”
That early tension between logic and artistry didn’t create friction, rather, it created fuel. Drawn into the world of animation and VFX, Yan quickly realized he could use his technical mastery to unlock new forms of storytelling. What started as a passion project became a career that bridged neuroscience, machine learning, animation and big-budget Hollywood productions.
Pipelines from scratch
From Disney Feature Animation to DreamWorks, Starz, and beyond, Yan has led some of the most ambitious studio builds in the business.
“One of the things I was always drawn to was greenfield opportunities,” he says. “I like building things from scratch – teams, pipelines, render farms, editorial workflows, all of it.”
That mindset proved essential when working on The Matrix Reloaded, a film that pushed the limits of early 2000s VFX. And later, when building out motion capture pipelines for Happy Feet 2, he found himself once again designing not just tools, but entirely new processes.
“When you’re creating a pipeline, it’s not just about technology, it’s about people and iteration. You’re enabling creativity through structure.”
GenAI, real-time, and the generalist
Yan is particularly excited about the current era, where real-time engines, procedural tools, and generative AI are not just disrupting pipelines but democratizing them.
“I think we’re in the middle of a renaissance for generalists,” he notes. “People who can code a bit, animate a bit, rig, design, train a model, these hybrid creatives are suddenly really valuable.”
Instead of treating AI as an existential threat or as hype fuel, Yan sees it as a practical extension of an artist’s toolkit. He emphasizes that successful studios will be the ones that embrace a “bimodal” approach, leveraging AI-native pipelines where appropriate, while continuing to value deep craft in other areas.
“The trick,” he says, “is knowing when to use AI and when to just let an artist paint.”
The architecture of innovation
At KMM, Yan’s current role as Studio Architect is all about building for flexibility. It’s not just about handling big VFX shots, it’s about futureproofing an entire creative ecosystem.
“The best architecture is invisible,” he says. “Artists shouldn’t have to think about the infrastructure. It should just work, like gravity.”
That principle has guided his recent work, integrating real-time engines like Unreal, AI-assisted asset generation, and new types of data orchestration into production workflows. But he’s cautious about over-engineering.
“You need just enough structure to support creativity, but not so much that it becomes rigid or fragile. The best pipelines evolve.”
Scaling creativity with infrastructure that delivers
As creative workflows become increasingly iterative and compute-heavy, the need for fast, scalable infrastructure is more critical than ever. For Yan Chen, the performance gains from using Dell and Nvidia-powered systems are not abstract, they’re measurable and transformative.
“We had 10 times more compute, so we needed 10 times less compute time. And we iterated 10 times more,” Yan says. “That’s almost a 100x increase. But yeah, 50x would be a conservative estimate.”
Dell PowerScale was the critical enabler behind those gains, delivering the fast, scalable storage needed to keep pace with 10× more compute. Massive scene files that once took 10 minutes (or more) to load now open in under 30 seconds, eliminating bottlenecks and empowering artists to iterate freely. With PowerScale’s modular, scale-out architecture, KMM could seamlessly grow performance and capacity as demands increased, supporting everything from real-time rendering to AI-assisted previsualization without compromise. Crucially, PowerScale feeds data fast enough to keep high-performance GPUs fully saturated, efficiently filling VRAM for AI and real-time workloads to perform at their peak.
That level of responsiveness turned idle GPU potential into real creative output. Assets moved faster, workflows stayed fluid, and infrastructure never got in the way. For Yan and his team, Dell’s storage layer wasn’t just fast – it was invisible, letting the focus shift from waiting on files to pushing creative.
It’s not just about raw power – it’s about empowering creatives to refine their vision more frequently and with less friction. The ability to deliver more iterations per production cycle dramatically improves both quality and cost efficiency.
“That allowed the director to see his film that many more times, to fine-tune it to his vision.”
When asked how to justify the investment in fast storage and high-end workstations, Yan breaks it into two paths:
“You show the visuals to the director, and you show the savings to the producer. If the director signs off the creative, and the producer signs off the cost and workflow, then you’re good.”
But the real shift, he says, comes from embracing a new creative model, where compute power is no longer just a support layer, but a core driver of the process.
“The model we’ve been using is probably half human labor and half compute. But the next step – the solo enterprise model, flips that. It’s 90% compute, 10% human.”
In that context, investments in infrastructure aren’t just justified -they’re essential. And for Yan, Dell’s scalable performance and reliability provide the creative runway that today’s storytelling demands.
AI isn’t coming. It’s already here.
When asked about the biggest shift in the industry, Yan is clear: “AI’s already here, it’s just not evenly distributed.”
He sees immense opportunity for studios willing to experiment and rethink their assumptions. But he also warns against treating AI as a silver bullet. Instead, he recommends focusing on well-defined use cases like background generation, asset tagging, or motion cleanup, where ROI is clear and the tools are mature.
“It’s like any other technology: know its strengths, respect its limitations, and integrate thoughtfully.”
AI as a new era
Yan doesn’t just see AI as a wave of innovation, he frames it as the beginning of a new civilizational age, akin to the geometric multi-age shift from the Stone Age to the Industrial Revolution. This framing moves the conversation far beyond productivity or efficiency into a broader, more philosophical view of how AI will alter every facet of human life, including storytelling and creative workflows.
“AI is not just a hype cycle. It is a new civilizational era… Everything is different now or soon will be different. Content creation is different. Coding is different. Chemical, biological and mechanical R&D are different. Every workflow is different. And the same goes for every other industry too.”
To underscore his point, Yan walks us through the compressed progression of human advancement – from the Stone Age to the Industrial Revolution to the Information Age, and now, the AI Age.
“If you go back 300,000 years ago, stone tools were everything. Then came the Bronze Age, then the Iron Age, then the Renaissance – each era changed everything. The Industrial Revolution lasted about 100 years, and then we hit the Information Age, which has been about 30 years. But AI? This new age will be far shorter and even more disruptive.”
That compression, he explains, is what makes the AI era so different and so urgent.
“If you take a natural geometric progression, the AI Age will last 5 years, maybe 10. But it’s going to be short, and it’s going to be different. Every workflow is different. Content creation is different. But so is finance, healthcare, logistics – every vertical is being transformed.”
To him, this isn’t a wave to surf, it’s a tectonic shift in how we build, create and think.
“That’s how I’m not understating how important of a moment this is in the history of mankind.”
Closing thoughts: build, adapt, repeat
Yan Chen’s career isn’t just a blueprint, it’s a lesson in evolution. From analog pipelines to AI-native workflows, he’s constantly adapting, constantly building. But at the core, one thing remains unchanged:
“It’s always about making things – tools, stories, teams – that work better together.”
In an industry often dominated by big names and louder voices, Yan is proof that the architects behind the curtain often shape the stage the rest of us perform on. His approach echoes a broader shift across the industry, toward integrated, data-centric foundations where infrastructure, AI and storage aren’t siloed but designed to move in sync. It’s a vision Dell shares deeply through its AI Data Platform, which aims to unify and accelerate the entire data journey so teams like Yan’s can focus less on the plumbing, and more on creating what’s next.
To learn more about how Dell Technologies helped power GenAI innovation for Kennedy Miller Mitchell, read the full story here.
