Rayd
Poster review, built on Claude

A second look.
Before the
final yes.

For teams that approve event posters on deadline. Rayd uses Claude to read the design and its approved brief side by side, then points reviewers to the wrong date, the missing line, or the unreadable text before it goes to print.

Request early access Try the sample review
One poster. Two perspectives.Sample / 01
A design gathering FUTURE
FORMS
24 OCT 2026

13:00 — 17:00
STUDIO 04

Fictional event / Original sample
Approved brief
Future Forms
Date25 Oct 2026
VenueStudio 04,
North Hall
ContactEvent desk
A date worth checking

24 Oct on the poster. 25 Oct in the brief.

Example
Original artwork and fictional event data, created for this demo.

The details connect.
The review should, too.

A poster can look ready while one critical detail is still wrong. Rayd brings the source, the design, and the reviewer into one review.

Information, against the brief

Compare the title, date, venue, and contact instructions with the approved source.

Visuals, in context

Surface readability, placement, and missing elements for a person to inspect.

Evidence, before approval

Keep each concern connected to a poster region and a source requirement. People make the final decision.

Who it's for

Teams that sign off
on posters every week.

Wherever one brief turns into many posters, banners, and social cards, the same checks are repeated by hand, often by the person with the least time left before print.

Event & PR agencies

Several client briefs at once, each with its own dates, venues, and approval chain.

Campus organizations

Student councils and clubs publishing frequent events with rotating volunteer designers.

Venues & cultural centers

Program schedules that change late, and posters that must change with them.

Public communication teams

Notices where a wrong date or missing contact line reaches thousands of residents.

Inside the sample review

Follow a finding
back to its source.

Select a concern to highlight its location and compare the evidence. Switch to the revised poster to see the corresponding example edit.

Sample walkthroughFixed illustrative findings. No AI analysis, upload, or automatic approval.
Future FormsFictional design event

Poster

Original sample
Future Forms / 2026FF

FUTURE
FORMS

Synthetic event artwork · Rayd demo

The outlined area is the selected evidence region. You can also select a region directly.

View the selected finding

Approved brief

Sample source
Event requirements

Use the approved details below on the final poster.

Title
Future Forms
Date & time
25 October 2026
13:00–17:00
Venue
Studio 04,
North Hall
Contact instruction
Questions: event desk
Visual requirement
Keep supporting copy clearly readable against the background.

Synthetic brief. All names, locations, and event details are fictional.

Illustrative findings

4

Approved source
Poster evidence

Local demo state only. This does not approve the poster.

Return to the poster
What makes a finding structured?

Each sample finding carries a category, source requirement, observed content, and normalized poster region. A review flag records only your choices in this browser session.


        

      

    


    

      
Designed around
a human decision.
The intended MCP-based workflow connects inputs, model-assisted checks, and evidence. It keeps final approval with the team reviewing the work.

      
Planned architectureNot connected in this demo

      
    
        
  1. 1
    Poster + approved brief
    Bring the design and its source requirements together.
    2. 
        
  2. 2
    Claude + MCP checks
    Claude reads both inputs and checks information and visual concerns.
    3. 
        
  3. 3
    Evidence-linked findings
    Return structured observations tied to the source and poster region.
    4. 
        
  4. 4
    People review & approve
    Inspect the evidence, resolve concerns, and decide what is ready.
    5. 
      

    


    

      

        
Why Claude
The job needs a model
that reads both sides.
Poster review is a multimodal comparison: an image on one side, a written brief on the other, and a structured answer people can act on. Each step maps to a Claude capability.

        

          
Vision
Read the poster as designed
Extract dates, venues, and small print directly from the artwork, and judge readability in context.

          
Documents
Hold the full brief in view
Load the approved brief, including PDFs and revision notes, as the source of truth for every check.

          
Structured output
Return findings, not prose
Each finding follows the JSON schema shown in the demo: category, source field, observed text, poster region.

          
MCP
Work where files already live
Connect to the storage and design tools teams already use, so review happens without re-uploading.

        

      

    


    

      
Benchmark
Measured on posters
with known errors.
We built a benchmark of synthetic posters with seeded errors and clean controls, then ask Claude to review each poster against its brief. The set, the prompt, and the scoring script are part of the project.

      

        
        

          

            
Synthetic posters
20

            
Seeded issues
21

            
Clean controls
4

            
Issue types
6

          

          
Title typos · wrong dates · wrong times · incomplete venues · missing contact lines · low-contrast text

          
          
Claude results for this set will be published here after the first full run.

        

      

    


    

      

        
Team & roadmap
A focused start.
A clear next step.
Rayd is built by a solo founder covering product and engineering, focused on one workflow: getting event posters approved without avoidable errors.
Contactteamsort@rayd.cc
Request early access

        

          
Done
Interactive review prototype
The review experience, evidence linking, and finding schema are designed and testable on this page.

          
Built
Claude benchmark harness
Each poster and brief goes to the Claude API (vision + structured output); findings are scored against the known errors.

          
Next
Live review, then pilot
Bring the same pipeline into the product with poster uploads, then pilot with event and design teams.

        

      

    

    

      
Questions we hear.

      

        
Is the demo running AI analysis?
The interactive demo uses prewritten findings for one fictional event, so the review experience can be tried without uploads. Claude's actual output is measured in the benchmark section.

        
Can I check my own poster here?
Not in this version, which uses only synthetic samples. Uploads arrive with live review, the next milestone.

        
Does Rayd approve posters automatically?
No. Rayd surfaces evidence; a person makes the final decision. It does not make legal infringement determinations.