UAE · ENBook a demo
Procure AIDesignAssistant

Prototyping Agent

Builds clickable prototypes for testing.

Prototyping AgentProduct Experience
Industry
Procure AI
Department
Design
Team
Product Experience
Kind
Assistant
Approver
Product Experience Manager
Code
DSN-03

It builds a clickable prototype of platform screens, such as raising a requisition from a punch-out catalogue, approving on mobile or submitting a bid in a reverse auction, wires the hotspots for each test task and checks every task reaches its success screen. It sets up the usability test with tasks and screener, and holds anything going to client users or suppliers for approval.

How it works

6 steps, and a person signs off before step 6.

  1. 01Looks it up

    Finds the record the request is about in the design workspace: the frames, components and existing prototype links for the flow named..

  2. 02Reads the request

    Works out what is being asked and what a complete answer needs.

  3. 03Checks the facts

    Checks figures, dates and rules against the source records, not from memory.

  4. 04Works it through

    Prepares the draft, match, recommendation or reply, with its sources.

  5. 05Asks a person

    Hands the work to Product Experience Manager with its reasoning before anything goes out.

  6. 06Acts

    Once approved, it acts in the user-testing platform: it creates the unmoderated test with the prototype link, tasks and screener, launched or left as a draft. and records what was done.

What it works with

Your knowledge, your systems, your people.

Knows

The prototyping and usability testing playbook: fidelity levels, task-writing guidance, screener templates for each platform role, session lengths, the synthetic test data set (fictitious suppliers, POs and prices) and the rules for testing with client users.

Reads from

the design workspace: the frames, components and existing prototype links for the flow named.

Acts in

the user-testing platform: it creates the unmoderated test with the prototype link, tasks and screener, launched or left as a draft.

Hands back

An object with screens (a list, each with screen and hotspots (a list of element and target_screen)), test_tasks (a list, each with task, success_screen and path), dead_ends (a list), fidelity, screener (a list of questions), estimated_minutes and launch (true or false).

Asks a person when

The test goes to client users, suppliers or a recruited panel, has more than 5 participants, or the prototype holds real client or supplier data.

ApproverProduct Experience Manager

Settings

Your limits, set by your managers.

SettingDefaultWhat it controls
Most tasks per session5A test with more tasks than this is split or trimmed.
Session length (minutes)20The tasks must fit in this time, at about four minutes a task.
Fidelity when none is askedmidLow is wireframes, mid is greyscale with real layout, high is final visuals.
May use real client or supplier dataOffWhen off, the prototype uses the synthetic data set only: fictitious suppliers, PO numbers, prices and bank details.
Launch without approval up to (participants)5An internal test with this many participants or fewer launches without the approver.

Limits are settings, not code: a manager changes them without a developer.

Rules it always follows

  • Every task has a success screen reachable by clicking from the start screen; list any dead end before the test is set up.
  • A test has at most 5 tasks and fits in 20 minutes; trim the lowest-value tasks first.
  • Build at the asked fidelity, or mid when none is asked.
  • Word tasks as the user's buying or approval goal, such as order ten laptops for a new team, never naming the button to press.
  • When off is false, every supplier, price, PO number and bank detail is from the synthetic data set; a real client's prices never appear, even to that client.
  • A test going to client users, suppliers or a recruited panel, or to more than 5 participants, needs the Product Experience Manager, and client users are recruited only through their engagement lead.

See the Prototyping Agent agent on your own work.

We’ll run it on a sample of your work and show you what your approver would see.

Prototyping Agent · Procure AI | Basinwright Industries