There is something undeniably magical about watching a modern software developer operating in a complete flow state. Fingers fly across the mechanical keys, typing natural language prompts or code fragments into advanced development environments like Codex, Claude, or Cursor. In response, a powerful, responsive creation interface unfolds instantaneously, powering a workflow that would have felt entirely unrecognizable—perhaps even alien-like—to any professional software engineer just two short years ago. The entire integrated development environment, or IDE, has been thoroughly torn down and painstakingly rebuilt from the ground up specifically for the age of artificial intelligence. As a result, software developers are no longer just working; they are flying.
Over that exact same two-year span, generative video models have raced ahead at a breathtaking pace, achieving milestones that once seemed decades away. Yet, unlike the software coding ecosystem, professional video creation largely remains stuck in the stone age. Its entire foundational toolchain is still built around the cumbersome, legacy mechanics of capturing, uploading, storing, and manipulating enormous, heavy media files.
Generative video has officially arrived at a profound threshold moment. The underlying models have rapidly grown expressive enough to effectively function as the new camera. The primary bottleneck holding back the industry is no longer the technical artistry of traditional animation, nor is it the prohibitive nine-figure budgets traditionally required for massive film crews, physical sets, and extensive visual effects. Instead, the true contemporary bottlenecks are human creativity and token inference. Despite these incredible technological leaps, video creators are not yet flying.
Without a proper, purpose-built AI-native tool suite—what many in the industry have eagerly anticipated as a "Cursor for video"—the talented professionals who make feature films, serialized television shows, and high-end commercials for a living remain stuck in a frustrating limbo. They are forced to awkwardly bolt modern model outputs onto archaic, legacy editing pipelines. In this fractured ecosystem, generation, creative review, and production tracking each live in separate, disconnected applications. Day-to-day workflows constantly buckle under the sheer, overwhelming volume of generated material. Crucially, a finished take often cannot be traced back to its original source for legal clearance and intellectual property tracking. It is a gaping problem that represents the single biggest roadblock on the industry’s path to truly embracing artificial intelligence.
A new platform called Preview has arrived to clear that critical block. Designed as an AI-native video creation and production platform built specifically for professional creators, Preview seamlessly combines a traditional video timeline with an infinite canvas for ideation into one collaborative, multiplayer workspace. Within this unified environment, creative teams can generate video assets using any underlying model while meticulously tracking every single character, location, and prop. The platform maintains complete, verifiable records of how every asset was made, enabling both halves of a hybrid production—traditional media production and AI generation—to work together smoothly and efficiently. There is simply nothing quite like it currently available on the market.
The people behind this ambitious vision are co-founders Stefan Fejes and Veljko Tornjanski, who have been best friends since their high school days in Serbia and seem to have been tailor-made for this exact mission. Their professional backgrounds map perfectly onto the challenges of the space. They were early founding team members at Veed, a pioneer in accessible video creation tools. Later, at Vizcom, Veljko served as the very first hire and eventually stepped up as Head of AI, while Stefan led the product organization. Through these experiences, they have developed a deep, nuanced understanding of the notoriously tricky mechanics of video production, coupled with a genuine empathy for their customers and a proven track record of delivering sophisticated, polished work on impressively short timelines.
That product-market alignment is already bearing fruit in the real world. More than 100 studios are actively using Preview in their daily operations, ranging from nimble creative agencies crafting high-end commercials for Fortune 500 brands to cutting-edge AI-native studios and established Hollywood feature filmmakers. Simply put, these early adopters love the platform. Customers regularly report that Preview has quickly become the most significant and relied-upon tool in their daily work, serving as the single source of truth from which all downstream inference flows.
The broader industry stands at the precipice of a massive video token tsunami. Over the coming years, many billions of dollars of inference will flow directly through this software category, driving the next wave of iconic films and valuable intellectual property. In this shifting landscape, maximum economic and creative value will naturally accrue to the place where the actual work happens—the platform that a creative team opens first thing in the morning and closes only when the workday is done, where critical decisions get made and the audit record gets kept. Preview is positioned to be that platform, backed by the right team, the right product, and arriving at precisely the right moment.
Comments (0)