AI Readiness
Prove your AI-built software is governed.
Enterprises no longer just ask whether you use AI — they ask how it’s governed. In the EU AI Act era, “we use AI responsibly” is an assertion. ShipReady measures how AI is used to build your software, from your own systems, so you can show it.
What ShipReady measures
AI ROI, AI Readiness, and Agent Health are deterministic graders, not an LLM — see How we use AI. ShipReady measures and helps you govern AI in your SDLC. It does not certify your AI program, and it never invents a score it cannot measure.
How the AI scores are computed
AI Readiness reads your connected systems, starting with your repositories, and looks at how AI shows up in the way you build: adoption, AI-assisted development, and the practices that keep it reviewable. The grader is deterministic. The same inputs always produce the same number, and every point on the score traces back to the signal that produced it.
When a signal cannot be read, the score says so. A repository that is out of scope, a source that is not connected, or a metric that needs data you have not provided all read as not measured. Nothing is filled in with an industry average, and nothing is estimated by a language model and presented as a measurement.
Answering the buyer's AI question
Security questionnaires and procurement reviews now ask how AI is used to write the software being bought, who reviews that code, and what policy governs it. A slide that says the team uses AI responsibly does not answer that. A score with its inputs, its coverage, and an honest list of what was not measured does.
ShipReady gives you that answer from your own systems. You can show a reviewer which signals were read, which were missing, and why a number is what it is. The same evidence feeds the rest of your ShipReady Score, so the AI story and the engineering story never contradict each other.
What you need to get a first number
Connect source control first. AI Readiness is scored from your repositories, so the first number arrives without a long setup project. AI ROI becomes measurable once you provide your own cost inputs and adoption is metered, and Agent Health once agent telemetry is connected. Until then both stay visibly not measured, which is useful in its own right: it shows a reviewer exactly what you have instrumented and what you have not.
Repository scope applies to every AI score. If some repositories are archived experiments or documentation, exclude them once and the scores respect that choice. The number you show a buyer then describes the software you actually ship, not everything that has ever lived in your account.
Frequently asked questions
- Does ShipReady certify our AI program?
- No. ShipReady measures how AI is used to build your software and helps you govern it. It does not certify your AI program, and no score is an audit opinion. Use it to show evidence in a review, not as a certificate.
- Why does Agent Health read not measured?
- Agent Health describes how reliably your coding agents execute, and that needs telemetry from the agents themselves. Until you connect agent telemetry, Agent Health reads not measured instead of showing a guessed number.
- Where do the AI ROI numbers come from?
- From your own cost inputs and your own adoption data. ShipReady does not substitute industry benchmarks. Until adoption is metered, AI ROI reads not measured.
- Is the AI Readiness score produced by a language model?
- No. AI Readiness, AI ROI, and Agent Health are deterministic graders. The same inputs produce the same score, and each point traces back to the signal behind it.
- How is AI Readiness different from AI governance?
- AI Readiness measures how AI is used in your engineering work. AI governance covers the AI systems you build and use as products and services. Both sit on the same evidence model, so a reviewer sees one consistent record.