Free tool

Score one workflow before you build.

Review seven dimensions of AI readiness, identify blockers and decide what evidence to gather next. Use the same workflow for every answer. No email required.

Choose one workflow and score the evidence you have today. No account or email is needed. Your selections stay in this page and reset when you reload it.

1: Blocker
A required foundation is missing or cannot yet be demonstrated.
2: Needs work
Some evidence exists, but the gap or exception handling is unresolved.
3: Pilot foundations
Documented and checked on representative examples for a limited pilot.
4: Operating evidence
Tested in the intended setting, with an owner, controls and ongoing review.

Can the workflow owner access representative inputs with the data owner's permission? Check approved exports, permissions and sample records.

Are the inputs accurate and complete enough for this task? Review missing fields, inconsistent records and difficult examples.

Is there a supported way to read inputs and return reviewed outputs? Confirm APIs or exports, permissions and a manual fallback.

Are the trigger, steps, exceptions and acceptable output documented? Bring an example of both a normal run and a failed one.

Who approves outputs, handles exceptions and maintains the workflow? Name an owner and a backup, with time to review the pilot.

Are data handling, access and human approvals agreed? Check what may leave your systems and which actions must wait for a person.

Can the team detect errors and stop or recover a failed run? Define review measures, an exception log and a fallback process.

0 of 7 dimensions scored

Finish all seven dimensions

Choose a score backed by evidence for each dimension. If a required foundation is missing, score it 1.

This is Cyndra's discussion rubric, not a standard, certification or safety approval. A single 1 is a blocker whatever the total. Review the evidence and applicable requirements with the responsible people.

How to read it

Turn the score into questions to resolve

A score of 1 in any dimension takes priority over the total. Record what is missing, who will resolve it and what evidence will show the gap is closed. A strong total cannot compensate for missing access, ownership or controls.

Without a blocker, totals up to 16 suggest work on the foundations. Scores from 17 to 23 support a conversation about a limited pilot; 24 to 28 indicate stronger foundations to review with the responsible owners. These bands are Cyndra's practical interpretation, not a validated standard.

Keep the evidence alongside the score: sample inputs, permission records, named reviewers, exception tests and a manual fallback. Agree acceptance checks before running the pilot and compare reviewed output, corrections and operating effort with the current process.

Read the AI readiness assessment guide for the full rubric, framework comparison and how to choose assessment support.

FAQ

AI readiness, answered

What does the AI readiness scorecard measure?

Seven dimensions for one workflow: data access, data quality, integration, workflow definition, ownership, security and controls, and monitoring. Score each from 1 to 4 using evidence from the task you want to improve.

Does a high score mean we can deploy?

No. This self-assessment is a discussion aid, not a certification or approval. A single score of 1 is a blocker regardless of the total. The responsible workflow, data and security owners must review the evidence and agree acceptance checks before deciding to proceed.

Do I need an account or email?

No. Select a score for all seven dimensions to see the total and next step. Selections are kept only in this page and reset on reload. Gathering the evidence behind your scores may take longer than completing the form.

What happens after I get my score?

Record the gaps, name an owner for each and agree the next test. You can bring one recurring task to Cyndra's free live build to examine inputs, review steps and implementation options with the team.

Next step

Turn your score
into a plan

Bring one recurring task to a free live build. Review its inputs, ownership and approval steps with the people who would build it.