Skip to content

Your ecommerce work.
One agent workspace.

Bring product knowledge, research, listings, and marketing work together.
Work with focused AI agents, then review and keep the results.

Open workspaceExplore capabilities

Prepare an Amazon listing

Help me turn these product details into an Amazon listing draft.

A clear plan, before the work

  1. 1Review product facts
  2. 2Research public references
  3. 3Draft a title and five bullets
  4. 4Prepare image directions

Review the plan before image generation or external changes.

Illustrative workflow · No live task is running

One workspace.
Many kinds of work.

Start with the task. Bring in the capabilities it needs.

01

Product & Listings

Organize product facts. Draft Amazon listing copy and image plans.

Product brief · Title & five bullets · Image plan02

Market Research

Review public sources, competitors, and customer needs.

Research notes · Source links · Reviewable findings03

Creator Collaboration

Find relevant creators, review their content, and build your shortlist.

Creator profiles · Candidate library · Outreach preparation04

Visual Creation

Generate images, split image layers, and reuse visual styles.

Brand visuals · Image layers · Style templates

Focused agents.
A shared way to work.

Give each agent a role. Add the skills, tools, and connected context your work needs.

Open workspace
Team
ProductListingResearchCreatorVisual
Capabilities

SkillsToolsConnectorsKnowledge

Work

Plans, sources, and deliverables in the same workspace.

  1. 01

    Describe the goal

    Tell the agent what you want to accomplish.

  2. 02

    Review the plan

    Check the approach, product facts, and sources.

  3. 03

    Keep the deliverables

    Refine and reuse the outputs in your workflow.

Confirm image generation and external changes before they run.

Go deeper when the task calls for it.

Open a focused app for creator collaboration or visual creation. Both sit within the wider ecommerce workspace.

More than an image.
A starting point.

Generate fresh visuals. Split them into image layers. Save a style you can return to.

Explore Visual Studio
Original image
Original image
Editable image layers
Editable image layers
Make it your own
Make it your own

Image Generation

Turn your ideas into new visual directions.

Split Layers

Separate images into elements you can rearrange.

Style Templates

Reuse a visual style across your brand.

Illustrative workflow. AI-separated elements are image layers; editable text and vector recovery are not guaranteed.

Built around the way ecommerce teams work.

01

Amazon sellers

Organize product facts, prepare listing copy, and review image plans before you use them.

Explore the workflow
02

Ecommerce operations teams

Research a market, compare product messaging, and keep useful sources close to your work.

Explore the workflow
03

Brand & marketing teams

Bring research, creator collaboration, and visual production into the same work context.

Explore the workflow

Keep creating.
Right in Figma.

A familiar home for your design process. Create in your browser or in Figma—your choice.

See the Figma workflow
  1. Open the Pavos plugin in Figma Desktop and sign in with your Pavos account.
  2. Select an image or upload one, then choose Split Layers.
  3. Review the result and add the separated images to a new Figma frame.
The existing Figma plugin remains a separate way to use Pavos.
Pavos plugin in Figma, showing image layer separation and a layer list.
Figma + Pavos

A few things, explained.

What is Pavos?

Pavos is an AI agent workspace for ecommerce teams. Organize product facts, research public sources, draft Amazon listings, and prepare marketing work with focused agents. Creator Hub and Visual Studio are specialist capabilities within that workspace.

What kind of ecommerce work can agents help with?

Start with a product brief, a research question, a listing draft, or a marketing task. Agents can organize facts, prepare source-backed research, draft copy and image plans, and help with creator or visual workflows. Available actions depend on configured services; review the plan and outputs before use.

Does creator discovery support Instagram, TikTok, and YouTube?

Yes. Creator discovery supports Instagram, TikTok, and YouTube, with platform, region, language, and category filters. Available data depends on connected sources. The website features curated public TikTok and YouTube profiles; changing filters here does not run a live search.

Can I use Visual Studio without creator tools?

Yes. Visual Studio and Creator Hub have separate entry points. You can focus on image creation in the browser or use the Pavos Figma plugin as part of your design process.

Are all separated layers editable text or vectors?

Separated elements are primarily image layers. You can rearrange, resize, hide, and reuse them, but lettering may stay inside an image. Review AI results against the original and add your own text when precise typography matters.

Can the workspace support new capabilities over time?

Agents have defined roles and can use configured skills, tools, connectors, and workspace knowledge. These can extend the kinds of tasks you bring into the workspace. Services and permissions must be configured before a connected action is available.

How do accounts and AI credits work?

The local website and Agent workspace share local accounts and credits. Demo tasks are free; authenticated model calls use actual token usage, and GPT Image 2 VIP generation costs 40 credits per image. Local accounts and test top-ups are not connected to Supabase or payments. Live Figma credits remain separate.

Your next ecommerce task starts here.

Bring your goal, your product context, and your team.

Open workspace