RequirementsDNA
Turn legislation into requirements you can trace.
RequirementsDNA turns an Act into user stories for the system that has to carry it out. Every provision ends up in a story or is excluded with a recorded reason. Each story traces back to its provision and keeps the Act’s own word, such as “must” or “may”.
Co-developed with ConceptSix (opens in a new tab). We run each engagement with your team and deliver the finished workbook.
How it works
It starts from the Act’s own words
We split the Act into one record for each subsection, following the Act’s own contents page. Each record keeps the exact words and the page they appear on.
Where a subsection lists separate things to build, we split it into one record per item. Where your practitioners have noted who acts under a section, we join that note to every record in it.
Written rules decide what becomes a story
AI agents group the subsections into user stories, following a written grouping protocol. A story covers one actor and one workflow, and a subsection is never split between stories.
The agents exclude a provision only for a set reason, such as a definition or a duty that belongs to someone else. Every decision records its reason, so a reviewer can check it.
The Act’s own word sets each criterion
The agents write each story from the provision text alone, under a second written protocol. Each story names the provisions behind it.
“Must” becomes something the system requires, “may” something it allows, and “should” a prompt. A discretion stays a discretion, so the system records the choice and the reason rather than forcing it.
Nothing is merged unchecked
Automated checks run each time new groupings or stories are merged, and work that fails is held back. They confirm that every provision has exactly one decision and every exclusion has a reason. They also confirm that no story requires a duty when no provision in it imposes one.
We review the full grouping plan before any story is written, then review every story. An independent audit then checks the work against the Act itself.
One workbook, linked both ways
You receive one Excel workbook. The Stories sheet holds each story with its acceptance criteria and the provisions behind it. The Provisions sheet lists every provision with its story or its reason.
Each story links to its main provision, and the Provisions sheet shows which story each provision went into. When the law changes, you can see which stories change with it.
Co-developed with ConceptSix
RequirementsDNA was co-developed with ConceptSix (opens in a new tab), a consultancy in business strategy and technology transformation.
ConceptSix brings business analysis and runs workshops with your team. We run the method and build the workbook.
Where it is used
Delivered for a state child protection department, with ConceptSix (opens in a new tab).
Security and data
- Your data
- Your data is not used to train models.
- What the agents work from
- The AI agents that group and write the stories work from the Act’s text, your practitioners’ notes and the written protocols.
- How we work
- We run RequirementsDNA with your team, together with ConceptSix (opens in a new tab). Your staff receive the workbook, with no software to install.
Try RequirementsDNA on the Act behind your system
We run a pilot with your team on your Act, alongside your current process.
Start a pilot