Copilot Ask
AVAILABLEControlled release · chart-aware · public recorded example.
AI SEATING CHART · CURRENT CAPABILITIES
An AI seating chart generator with review and publishing controls.
Copilot answers questions about the chart open in the Designer, changes it when you ask, and can take a plan or reference image as input. Every change arrives as a revision you review before it publishes. AI-led 3D improvement is not part of this release.
SeatLayer’s AI seating chart generator is Copilot, working inside the venue designer: it answers questions about the open chart, edits it on request, creates a chart from written requirements, and takes a plan or reference image as input. Every AI revision is reviewed before it publishes.
CURRENT MATURITY REVIEWED 16 AUG 2026
New conversation
RELEASEDClear a long chat and begin with fresh conversation context.
Designer MCP
AVAILABLECompatible MCP clients · OAuth · one authorized chart scope.
Create and Edit
AVAILABLEControlled release · Copilot changes the open chart on request; you review the revision before it publishes.
NO SILENT PUBLISHCANONICAL API AUTHORITYHUMAN REVIEWVISIBLE COST BOUNDARY2D FALLBACK
01 · REFERENCEStart with a private venue reference
A plan or venue image supplies context. It remains a reference—not automatic venue truth.
WHAT AI ACTUALLY SEES
Create a seating chart from a venue image
The useful foundation is the chart model: rows, seats, tables, sections, labels, categories, accessibility data, floors, focal points, and published revisions. Ask receives an authorized summary of the open chart; it does not become the authority for chart state.
A venue image enters that model as a reference, not as a tracing: you can give Copilot a plan or reference image as input to the chart it is working on, uncertainty is disclosed, traces stay private, and your venue data remains authoritative. Arbitrary image or CAD fidelity is not promised — there is no claim of universal photo understanding, vector round-tripping, or automatic venue truth. If you would rather place the layout yourself, the free seating chart maker is a released route with no AI in the loop.
ONE CONTROLLED PATH
Natural-language commands run within product controls.
The model may interpret intent and propose a plan. Authorization, tool execution, canonical revisions, validation, evidence, billing, and publish confirmation stay outside the model.
COPILOT SESSION · CREATE AND EDIT AVAILABLEILLUSTRATED EXAMPLE
YOUAdd two rows at the back of the Balcony, same spacing as the row in front.
COPILOTPlanned two rows behind Balcony U, matching its spacing and curve. Capacity check: 1,240 → 1,288. Accessibility check: the four wheelchair spaces on Balcony A are unchanged. Prepared as a revision for you to review — nothing is published.
REVISION · NOT PUBLISHED+2 rows · Balcony V, W48 seats · capacity 1,240 → 1,288
Review changesPublishNO SILENT PUBLISH · YOU PUBLISH
An illustration of the review gate in the site's own design language — not a screenshot, a recorded session, or a live control. Copilot proposes; the revision, the validation result and the publish decision stay with you.
User asks or approves.
Every run begins in an authenticated workspace and an explicit chart scope.
Product resolves what is allowed.
Unavailable tools remain unavailable even if a model requests or claims approval.
API owns state changes.
Idempotency, revision conflicts, billing records, and confirmation come from canonical records.
User sees what changed.
Validation, before/after evidence, rollback scope, and publish state complete the result.
WHAT COPILOT CAN DO
Start with the job you need done.
See which inputs Copilot accepts, what it can change, and what you review before publishing.
Ask about a chart
Explain capacity, sections, accessibility, labels, and review priorities without changing chart state.
Edit an existing chart
Bounded tools and revision-safe writes: a change lands as a reviewable revision with conflict handling and full-scope rollback, and publishing stays a separate step you take.
Create from requirements
Describe the room and Copilot plans the geometry, runs capacity and accessibility checks, and hands you a chart to approve. Human approval stays required before anything publishes.
Build from a reference
Give Copilot a plan or reference image as input to the chart it is working on. Uncertainty is disclosed, traces stay private, and your venue data remains authoritative.
Improve 3D · Off
Requires actual WebGL before/after proof on the same revision, supported-property changes, mobile checks, and a complete 2D fallback.
Work through an external agent
Claude, Codex, or another compatible MCP client can use one authorized chart scope. SeatLayer retains write and publish authority.
Help a buyer find seats
Buyer WebMCP is a separate opt-in browser workflow: a compatible assistant can read an event, find and select seats on the real map, inspect the selection, and optionally place a hold. It cannot check out or take payment.
WAYS TO START
Start blank, from a template, or with a supported reference.
Use a SeatLayer template, import a supported public seats.io chart into a draft you review, or provide a private source plan through the Designer. Orders, tickets, payments, customers, and private vendor data do not move with chart geometry.
Blank and template starts
Use the visual Designer and retain ordinary editable chart objects.
Public seats.io import
Create a native review-required draft; never auto-publish a cutover.
Private reference workflow
Authorize one chart and keep trace material private from buyers.
Arbitrary image or CAD fidelity
No claim of universal photo understanding, vector round-tripping, or automatic venue truth.
COMMON QUESTIONS
What users can rely on now.
Can Copilot create my whole chart?
Creating a chart from written requirements is an available capability, as is taking a plan or reference image as input to the chart Copilot is working on. Copilot plans the geometry, runs capacity and accessibility checks, and hands you a chart to approve — human approval stays required, and nothing publishes on its own. The limits are unchanged: arbitrary image or CAD fidelity is not promised, there is no claim of universal photo understanding, vector round-tripping, or automatic venue truth, and AI-led 3D improvement is off.
Can Copilot edit height or make the chart 3D?
No. The released 3D buyer view is separate from AI-led 3D improvement. Height and related properties remain normal Designer controls.
Can I connect Claude, Codex, or ChatGPT?
Designer MCP is a chart-scoped capability for compatible remote-MCP clients. Client support and setup differ; SeatLayer’s docs show each supported route and preserve OAuth and human publish boundaries.
Will AI usage consume sold-seat credits?
The public estimator covers confirmed sold seats only. SeatLayer has not published an AI allowance or AI price. Any applicable entitlement and cost boundary must be shown before use.
Where is the current source of truth?
The machine-readable capability registry is published at /ai-capabilities.json. Product and marketing surfaces are checked against that registry.
What is an AI seating chart generator?
It is software that turns a question, a written requirement, or a reference into work on a structured seating chart rather than a picture of one. In SeatLayer that is Copilot inside the Venue Designer: it answers questions about the open chart, edits it on request, creates a chart from written requirements, and takes a plan or reference image as input. Rows, seats, tables, sections, labels, categories, accessibility data, floors and published revisions stay the authority, and the API owns every state change.
Can AI create a seating chart from a venue image?
Yes — as a reference rather than a tracing. You give Copilot a plan or reference image as input to the chart it is working on; uncertainty is disclosed, traces stay private, and your venue data remains authoritative. Arbitrary image or CAD fidelity is not promised: there is no claim of universal photo understanding, vector round-tripping, or automatic venue truth, and what comes back is a revision you review before it publishes.
Can AI generate a seating chart from a floor plan photo?
A floor plan is one kind of plan or reference image the same workflow accepts as input. The released source routes are a blank start, a SeatLayer template, a supported public seats.io import into a review-required native draft, or a private reference workflow through current Designer and MCP boundaries — so a photo informs the draft you then check, it does not become automatic venue truth.
Does the AI publish chart changes automatically?
No — there is no silent publish. A change lands as a reviewable revision with conflict handling and full-scope rollback, and publishing stays a separate step you take. Authorization, tool execution, canonical revisions, validation, evidence and publish confirmation all stay outside the model, and unavailable tools remain unavailable even if a model requests or claims approval. Current states and authority boundaries are documented in the agent capability matrix.
Start with the chart and the currently released capability.
Use the released maker for a quick layout, Copilot Ask for chart understanding, or the agent documentation for a controlled integration workflow.