Hire for Trios: Running an AI-Native Studio
I just read FBRC.ai’s AI x Media & Entertainment — Studio Edition (Q1 2025) by Rachel Joy Victor, Todd Terrazas, and Eric Wilker. It’s a Q1 snapshot that hit in early Q2; reading it in August, you can feel how quickly the ground keeps shifting.
What doesn’t change is the signal: this is not a story about which AI tools you are using, it’s an operating-model story.
The report shows a field led by tiny, bootstrapped teams (roughly 70% of AI studios are five people or fewer) and a continuum where “AI-native” and “traditional with AI” already blur. That’s the world we operate in at Anima.
Distribution of Team Size among 98 AI-studios interviewed. Source: AI x Media & Entertainment — Studio Edition (Q1 2025) by FBRC.ai
We built Anima to be small by design—four founders, bootstrapped—because small is the only way to move at the speed of models, clients, and culture.
The question I obsess over isn’t “which model is best?” It’s: how do we structure people and pipelines so story survives model churn?
The studios in the report that are pulling ahead make their edge in the middle of the stack: model-agnostic pipelines, node-based control, and proprietary look libraries trained on their own work. That’s where art meets ops, and where a studio becomes a system.
Our core unit: a trio, not a department
Right now we’re four co-founders; when we scale, we’ll add trios—not departments. A trio is the smallest stable cell that balances taste, systems, and craft:
Director of Taste — owns world rules, shot intent, and final yes/no.
Pipeline Architect — composes node graphs, manages model routing, enforces “swap-safety” (change a model, keep the look).
Craft Specialist — motion/comp/3D/animation; turns keyframes into finals.
What a day feels like inside the trio: mornings set story and frames (we lock intent, not tooling); mid-day is pipeline surgery (swap models without drifting the look); afternoons are polish (motion, continuity, surface detail). This mirrors a broader pattern in the report: pre-viz frames increasingly become final frames—production and post become “polishing pre.”
When fidelity climbs, we snap in specialists—exactly the evolution the report tracks, from “one-man bands” to focused rosters that close the gap between out-of-the-box and professional fidelity.
The stack that lets a small team feel big
Our stack is intentionally boring. We mix closed providers with open ecosystems and route everything through Comfy-style node graphs, because surgical control beats prompt luck.
We train LoRAs per IP and per character, and we version “render recipes” like software.
Inputs: story bible, shot lists, visual grammar (lenses, lighting, grain).
Generation: model-agnostic; open + closed, chosen per scene goal.
Control: node graphs (ComfyUI), ControlNets, LoRAs/IP adapters; versioned recipes.
Refinement: video-to-video, style transfer, grey-boxed environments, comp/VFX overlays.
Infra: containerized micro-services, lineage/audit trails so decisions are trackable at scale.
The Tooling Stack of an AI-Native Creative Studio - Source: AI x Media & Entertainment — Studio Edition (Q1 2025) by FBRC.ai
The report is clear: most real innovation is happening in the middle layer; studios differentiate with workflow composition and datasets more than model worship. And as teams move from tests to production, containerized microservice backbones matter.
On sensitive brand work, we adopt a clean-model posture wherever possible—because safety and clarity beat novelty when you’re protecting IP. The report flags Asteria/Moonvalley’s Marey as a signal of where enterprise-grade pipelines are headed.
Why “small by design” is a feature, not a constraint
Speed of iteration: a few people can now do what once took dozens—fewer hops between intent and image.
Taste survives the tooling: when models change weekly, a trio guarding story, look, and pipeline keeps outputs coherent.
The boundary is gone: generative can be the shot or the scaffolding around live action and 3D; audiences care about outcomes, not labels.
If Q1 captured the end of “press-a-button cinema,” Q3 is the rise of the pipeline-native studio: small, narrative-led, stack-smart. At Anima, “hire for trios” is how we keep that promise honest—three people who can ship, supported by a system that keeps our look stable while the tools keep changing.
I highly recommend reading the full report to get a perfect snapshot of the current state of AI-native Creative Studios. You can find it here: https://www.fbrc.ai/reports
About the series & the author
This article is part of an ongoing series exploring the monumental shift of generative AI across the creative industries—from business and philosophy to technology and storytelling. We’re challenging myths, unpacking value, and redefining what creativity means in this new era.
I’m Riccardo Fredro, founder of Anima Studios—where we fuse cutting‑edge AI with over a decade of storytelling, filmmaking, marketing, and sound design to craft visually stunning, emotionally resonant content. Through this series, I’ll share insights and actionable perspectives on how generative AI is transforming leadership, culture, and the creative economy.
Stay tuned for more — and let’s reimagine creativity together.
