Every AI video model currently on the market produces the same thing: a flat file. You can watch it, share it, and trim it at the edges, but if you want to change the font, adjust an animation, or swap the background, you are essentially starting again.
That constraint has kept AI video at arm's length from professional production workflows, where video has always been built in layers, separate tracks for text, motion graphics, footage, and audio that can be adjusted independently until the moment of final render.
Idomoo, the Israeli enterprise video personalisation company, is today announcing Strata, a generative AI foundation model it claims is the first purpose-built for layered video output. Rather than generating pixels, as Danny Kalish, Idomoo's co-founder and CTO, frames it, Strata generates structure: independent layers with typography, animation, motion paths, and synchronised audio, all composed into what the company calls a “production-ready video blueprint.”
The distinction is architectural. Standard diffusion models collapse everything into a single tensor during generation; the spatial and temporal relationships between elements are baked into the pixels themselves.
Strata, according to Idomoo, solves a different computational problem: it designs the full composition, defining placement, contrast, movement, timing, and pacing across all layers simultaneously, while also enforcing brand guidelines. The resulting output is editable at the layer level, in the same way a professional would work in Adobe After Effects.
Strata is part of Lucas, Idomoo's AI video agent, which sits on top of the company's existing Next Generation Video Platform. One of its more technically specific capabilities is brand awareness: Lucas analyses a company's approved content to extract what Idomoo calls Brand DNA, covering design, narrative, and assets.
Strata then applies that specification to every video generated through the platform, enforcing typography, motion cues, colour values, and tone of voice across output at scale. The intent is to eliminate the template workaround that most “AI wrapper” products currently rely on, where generated footage is constrained to fit preset layouts.
Idomoo's argument is that forcing content into templates produces a recognisable visual compromise; Strata designs custom blueprints for each video instead.
Personalisation is the other axis. Because Strata's output is layered rather than flat, individual data fields, names, account details, transaction histories, product images, can be injected into specific layers of the video composition in real time.
This is the core of Idomoo's existing business model: the company's platform already serves JPMorganChase, Verizon, and American Airlines, among others, generating personalised videos at scale for customer communications, onboarding, and marketing.
Strata, in theory, makes that personalisation significantly more sophisticated, because it operates at the composition level rather than as an overlay on top of a pre-rendered clip.
The company is being careful about the launch's scope. An early access version is currently being tested by several of its largest customers. It is available now through Lucas AI Video Agent, but Idomoo has not disclosed which customers are in the early access cohort, what benchmarks the model has been tested against, or how it compares to off-the-shelf diffusion models on quality metrics.
The “first foundation model purpose-built for layered video” claim is Idomoo's own framing and has not been independently assessed. Strata technology is patent pending.
It is worth noting that Idomoo's own platform documentation, until now, stated explicitly that it used off-the-shelf foundation AI models rather than proprietary ones. Strata represents a significant shift in that positioning: from a company that applies AI to video to one that is building foundational AI for video.
Whether the underlying architecture delivers on that framing will become clearer as enterprise customers move from early access to production deployments.
Founded in 2007 by Yaron Kalish, Danny Kalish, and Assaf Fogel, the Ra'anana-based company has raised $27 million in total funding, including a $9 million Series A in 2013 and an $18 million Series B in 2019. It has been building its personalised video platform for enterprise clients for nearly two decades, which gives it an unusual amount of structured video production data to train on.