01 / STEP
Drop in your screens
Upload screenshots of any web or mobile flow, from a live product, staging, or a work-in-progress design. Or paste a public URL and ditx captures desktop and mobile views for you.
AI UX audits, grounded in evidence
AI UX audits that show what your users feel.
Get an AI UX audit of your product screens in minutes: a scored report with prioritized fixes, in English or Arabic.
Built for Arabic and right-to-left products, as well as English.
Free audit · No card needed · Results in 2–5 minutes
A quick first look
Try it now: paste your website URL
Private by default. We capture one screenshot of the public page you enter and delete it after 7 days unless you create an account.
Upload your product screens for a complete, evidence-based audit with prioritised fixes.
Start free auditWHAT YOU GET
Sample report shown in English and Arabic.
Same audit. Same evidence. Your team's language.
8Scored UX dimensions
100%Findings tied to a real screen
5Screens per Free audit · paid caps vary by plan
The difference is in the detail
An opinion is easy.
Evidence travels.
ditx keeps the chain intact: the screen the issue came from, the evidence behind it, the principle it relates to, and what the team can do next.
Screen
Signal
Finding
Decision
How it works
From upload to action in three moves.
No setup, no tracking scripts. Bring the screens you already have and get a plan your team can pick up tomorrow.
02 / STEP
ditx gathers evidence
Every screen is checked against usability heuristics and WCAG 2.2. Each finding links to its screen and cites the principle it relates to. Text contrast is measured from the pixels.
03 / STEP
Ship the right fixes
Findings are ranked by severity and mapped by impact and effort. Re-audit after you ship to see how the score changed.
Evidence-first reporting
No recommendation without proof.
Each finding pairs the problem with its evidence, the principle it relates to, and a recommendation, so your team can see why a change matters.
Finding #07 · Checkout, step 2
High severityPrimary CTA appears disabled
- Observed evidence
- Contrast ratio 1.4:1 between the button and its background.
- UX principle
- Visual hierarchy / affordance
- WCAG criterion
- 1.4.3 Contrast (Minimum)
- Recommendation
- Increase contrast and clearly differentiate active and disabled states.
- Assessment
- Contrast measured by code from the screenshot pixels.
Continue
Measured contrast1.4:1
Observed4.5:1 required
Download the full report as PDF or DOCX to share with your team.
Title
Evidence
Principle
Recommendation
Business impact
One review, eight lenses
See usability, accessibility, and flow in one view.
The overall score reflects eight scored UX dimensions, alongside a severity map showing which screens and categories need the most attention.
01Usability 58
02Accessibility 44
03Clarity 66
04Navigation 72
05Trust 70
06Forms 52
07Content 63
08Conversion 61
Example scores for the sample checkout above.
Priority matrix / roadmap
Know what to improve next.
Findings are ranked by severity, and an impact/effort matrix shows which ones are quick wins. Hover a dot to see the finding. The roadmap then groups the work into sprints.
Impact / effort matrix
Impact ↑
Effort →
Quick wins
Major projects
Fill-ins
Thankless tasks
Hover or tap a dot to see the finding.
Fix, then measure
Leave with a hand-off, not just a diagnosis.
Findings come with recommendations, engineering notes, and acceptance criteria where available. After the fixes ship, run another audit and see how the score changed.
Developer handoff · Finding #07
A brief engineering can act on, not a complaint list.
Each finding names the affected component, with the evidence, a recommendation, engineering notes and, where available, acceptance criteria.
- Finding
- Primary CTA appears disabled
- Affected area
- Checkout
- Component
- Continue button
- Evidence
- Button text contrast measured at 1.4:1 against its background.
- Recommendation
- Raise contrast to at least 4.5:1 and keep a distinct disabled state.
- Engineering notes
- Update the primary button color token, then check hover and focus states.
- Acceptance criteria
- Enabled button text reaches at least 4.5:1 contrast.
PDFDOCXExports include each finding's evidence, business impact and recommendation.
UX progress · Checkout flow
Ran an audit. Fixed issues. Ran another. Improved.
Each new audit round shows how the overall score changed from the one before.
- Round 386Improved +12 pts
- Round 274Improved +12 pts
- Round 162First audit
Example rounds. When two audits cover different screens, ditx says the scores aren't directly comparable.
Where clarity changes the work
Turn the hard-to-see into the next thing you fix.
For the team that owns the experience, whether they are shipping a first version or improving a mature product.
01 / Product teams
One shared source of truth for what needs to change.
Bring product, design, and engineering into the same evidence, then prioritize work without arguing from memory.
02 / Agencies
Make the why visible to every client.
Share the report with your client and show the evidence behind every recommendation.
03 / UX leaders
Prove the experience moved forward.
Re-audit after release and bring the score change into your next planning conversation.
Privacy and sensitive-data protection
ID details are masked before analysis.
When you upload a screenshot, an AI privacy check looks for government ID details and permanently masks them before the UX analysis runs. If a screenshot can't be checked safely, it's held back.
Example: a screenshot that shows an ID
Uploaded screen · Profile3 areas masked
National ID9861 042 337
Passport no.N 4821 337
Address14 Rainbow St, Amman
PlanTeam · Monthly
What the check looks for
National ID numbersPassport numbersTax and government IDsNames on ID documentsID photosDates and places of birthAddressesSignaturesQR codesBarcodes
What happens to every screenshot
- 1Checked on upload. An AI model looks for ID details on the screenshot.
- 2Masked for good. Detected areas are covered with a solid block. Blur isn't used, because blurred digits can sometimes be read.
- 3Confirmed by you. The audit can't start until you confirm your screenshots are ready.
"I have reviewed the screenshots for this audit and removed any sensitive information I do not want sent for AI analysis."
Required before every audit.
What an audit is based on
We say exactly what evidence a finding is based on.
Audits use screenshots only, and the report never claims to have tested behavior it couldn't see.
Screenshot audits
Findings are based on what's visible in the screenshots.
URL capture
Paste a public URL and ditx captures three desktop views and two mobile views to audit.
Human expert trust
AI gives you speed.
Experts give you confidence.
Every audit starts with AI evidence. When you want a human perspective, request a conversation with a UX professional: a review of your audit, or a workshop-style discussion with your team.
Topics you can ask about
- A human perspective on AI findings
- Questions about priorities and context
- A workshop-style team discussion
Included
AI Audit
Every finding is generated by AI and backed by visible evidence from your screens.
Optional
Expert review
Tell us what you need help with. We'll discuss whether a human review of a completed audit or a workshop-style conversation is the right fit.
Discuss expert support
Expert reviews are led by Hazem Hijazi, founder of ditx, with 20 years in design and a focus on English, Arabic and right-to-left products, supported by a growing network of experienced UX practitioners.
Founding testers
Be one of our first 20 founding testers
Get a free full audit plus a 30-minute review with Hazem, in exchange for honest feedback.
Why ditx exists
Hazem Hijazi
Founder of ditx
I've worked in design since 2006: graphic design first, then web, art direction, and UX for large digital products. During my career, I've worked on banking, fintech and government projects, which taught me that every screen has a business cost when it confuses people.
Across Arabic and English projects, I saw great teams ship interfaces that broke in right-to-left, failed basic accessibility, or buried the action users came for. Expert audits caught it, but they were slow and expensive. ditx is my answer: evidence-based audits in minutes, with a human expert there when you need one.
FAQ
Questions worth answering directly.
No. Every audit runs on AI. You can request a conversation about human review or a workshop; submitting a request does not authorize report access, payment, or a booking.
When you upload, an AI privacy check looks for government ID details, such as ID and passport numbers, and permanently masks them before the UX analysis runs. Emails, phone numbers and card details aren't detected automatically. Remove them before you upload. You confirm your screenshots are ready before every audit.
No. Audits use screenshots only. Findings are based on what's visible. Checks that need interaction, such as keyboard navigation, are listed as not evaluated, with the reason. Prototype coming soon; Staging environment coming soon; TestFlight coming soon; Interactive app access coming soon.
Only if you share it. Reports stay private to your workspace until you create a share link. A link can be open, password-protected, or limited to named emails, and every link expires within 5 working days. Company accounts can only share with verified emails on their own domain.
We don't claim a report is "100% secure" or "legally compliant". We document the specific controls ditx has built, so you can evaluate them against your own requirements.
ditx is in public beta and you can start on the free plan. Paid plans are being prepared and can't be bought yet.
Make the next move with proof
Less debate.
More direction.
Put your screens in view. Let the evidence show your team what deserves attention next.
Start free audit