01 · 20% of score
Business intent and requirements
Is work defined with testable acceptance criteria before anyone builds it, and are the decisions made along the way written down for the business owner to confirm?
Free assessment
Is AI helping your team ship better software, or just more code?
An honest look at whether your engineering organization turns AI capacity into reliable delivery.
Fourteen questions about how your organization defines work, builds with AI, proves the result and measures what changed. You get a profile across six capabilities, the gaps that matter most and practical next steps.
- 14 questions, about 7 minutes
- Results on screen right away
- No email, no account
Answers are stored without contact details to improve the assessment. If you request a second opinion, your email and note are linked to your results. The privacy page explains assessment records and ordinary server logs. Privacy
JavaScript is needed to take the assessment. If it does not load, refresh the page. You can still read about the assessment and the sample results below.
Sample results
Your reported AI workflow is ahead of verification.
Where you are
Developing
The weighted answers sit in the developing range. Individual capabilities can be stronger or weaker.
- Exploring
- Developing
- Systematizing
- AI native
Indicative 40 of 100, weighted across the six capabilities. Stages are a guide, not a scientific classification.
Six capabilities
Strong 60 and above · Partial 40 to 59 · Gap under 40 · Unknown has no score
Risky combinations
- Heavy AI use with weak verification
- AI adoption without better outcomes
Why it matters
AI amplifies whatever process it lands in.
Almost every team has coding agents now, so using them says little on its own. What decides the outcome is everything around the code. Clear requirements, real verification, sensible controls and honest measurement turn faster coding into better delivery. Without them, the same speed produces rework.
What it covers
Six capabilities, weighted by what it costs you when one is missing.
02 · 20% of score
AI native engineering workflow
How much of the work AI does, whether the team shares one way of working with it, and how far a large change can run on its own.
03 · 25% of score
Verification and software quality
Do the tests prove each acceptance criterion, and what must a change pass before it goes live? Weighted highest, because speed without proof is where the cost lands later.
04 · 10% of score
Engineering foundations
Can an agent read what an existing part of the system is supposed to do, and does every change reach production by the same automated path?
05 · 10% of score
Governance and accountability
Does everyone use approved company AI accounts, kept away from production, and is the ask written down and approved so engineering can be held to it?
06 · 15% of score
Delivery outcomes
Does work reach users faster, is it a good product, and does it ship without breaking things? Asked about directly, because activity counts can’t show it.
How it works
Answer for how things work today, not the plan.
Step 1
Answer fourteen questions
Each answer describes a behavior, not a grade. “I don’t know” is always an option, and it counts as unknown rather than zero.
Step 2
See your profile
A profile across six capabilities and where you are on the path from Exploring to AI native. Risky combinations are flagged whatever the average says.
Step 3
Take the next steps
Your three biggest opportunities, a 30, 60 and 90 day plan, and reading picked for your results. Useful whether or not you ever talk to anyone.
What to expect
A structured self evaluation, and nothing more.
What it is
Built on how Enso Dynamics plans, builds and verifies software with AI on real client work.
Method independent. A team with clear requirements and reliable verification scores well whatever process it follows.
Answers are stored without contact details to improve the assessment. If you request a second opinion, your email and note are linked to your results. The privacy page explains assessment records and ordinary server logs. Privacy
What it isn’t
- An audit. It reflects what you report, which can differ from what an outside review would find.
- A benchmark. There is no industry percentile, because there is no honest data for one yet.
- A tool quiz. Which AI tools you use matters far less than the process around them.