Light
Scale
Continuity
Style
Realism
Taste
Repeatability
AI Assisted Timelapse
A visual workflow for stabilising, extending and shaping long-duration image sequences across time, weather and change.
AI generates from probability.We shape it with control.
AI tends toward the average.We bring specificity, taste and judgment.
AI creates visual possibility.We resolve it into work that feels intentional, believable and finished.
Unnisbis builds controlled AI workflows and visual production systems for teams working with images, video, spaces, fashion, archives, social content and automation.
Traditional visual tools are deterministic. You move a layer, change a value, adjust a frame, and the result follows.
Generative AI works differently. A prompt does not execute a fixed instruction. It samples a possibility.
That is why an image can look impressive at first glance and still not be the image you meant.
The work is to close the distance between intention and output, with control.
Where the difference shows up
Unnisbis comes from "unnis bis ka farak", a Hindi expression for a subtle difference that still changes the result.
It is the difference between close and correct.
Between a promising generation and an image that keeps holding under attention.
Between something that looks good for a second and something that can be used, repeated and trusted.
That subtle difference is where we work.
Probability into intention.
How we help
Who it is for
Selected work
See all workSelected AI visual solutions that can run as standalone services, repeatable workflows, or be integrated into a client's existing production process.
A visual workflow for stabilising, extending and shaping long-duration image sequences across time, weather and change.
Research-led AI reconstruction workflows for damaged photographs, colour recovery and lost visual information.
Furnishing and styling empty interiors digitally — real furniture, light and placement set into the architecture exactly as photographed.
Controlled AI production workflows for campaign images, video scenes and branded visual outputs.
A repeatable visual content workflow for creating, testing and refining brand images, reels and campaign variants.
A brand-specific catalog workflow for creating styled product and fashion imagery with control over garments, bodies, looks and visual language.
Generative, algorithmic and projection-based image work exploring networked images, computation and new visual media.
Client-specific AI workflow solutions built around how teams already produce, review, approve and deliver work.
Founded by
Photographic judgment applied to generative systems.
Fabien Charuau brings together photography, coding, generative art, writing, research and more than two decades of professional image-making.
His work moves between visual culture and technical systems: architecture and interior photography, post-production, image analysis, generative experiments, writing on photography and teaching around AI.
Unnisbis brings that experience to bear on AI, turning possibility into controlled, intentional visual work.
Learn more about FabienBring us a visual problem, workflow, archive or idea that standard AI tools cannot resolve cleanly.
Start a conversation