Vizzy overview
Vizzy is Visualizee's agentic AI assistant. It reads your message and images, picks the right workflow, sets parameters, and generates — so you never have to choose a mode.
Vizzy is the assistant you talk to in every project. Older AI rendering tools make you pick a mode first — Inspiration, Render, or Motion — then manage settings yourself. Vizzy removes that friction: it figures out what you're trying to do and does it.
What "agentic" means here
On every message, Vizzy analyzes three things — what you said, what images you attached, and the conversation so far — and then decides the workflow on its own:
| You do this | Vizzy chooses |
|---|---|
| Describe a scene with no image | Inspiration (text-to-image) |
| Upload a sketch, "make it photorealistic" | Render (image-to-image) |
| "Animate this building" | Motion (image-to-video) |
| Upload 12 photos, "make them all watercolor" | Batch (multi-image) |
You never select any of that. You describe the outcome; Vizzy routes it.
How a turn works
Each time you send a message, Vizzy runs a short, consistent sequence:
Detect persona
Vizzy picks the one discipline that best fits your request — architect, interior designer, product photographer, and so on — and loads that persona's skill pack of rendering know-how.
Read intent
It classifies what you want: create something new, refine a small detail on a previous render, restyle the same subject in a new aesthetic, animate it, or explore several directions when your brief is still open.
Set parameters and generate
Vizzy chooses sensible defaults (aspect ratio, model, resolution, and for video, duration) and runs the generation. Results stream back into the chat as cards.
You're always in control
Agentic doesn't mean opaque. Two principles keep you in charge:
- Your manual choices win. If you set the aspect ratio or pick a model by hand, Vizzy respects it over its own default.
- Every change is visible. When Vizzy adjusts a setting, a small inline confirmation appears in the chat (for example, "Aspect ratio → 16:9" or "Switched to Motion · 5s"). Nothing happens silently.
When your intent is genuinely ambiguous — you upload images with no instructions, or it's unclear which images are references versus subjects — Vizzy asks a quick clarifying question instead of guessing wrong.