Skip to content

The passage

Four steps from
regulation to
engineering work

Landfall is a compliance-to-engineering tool for public bodies, banks and product teams: it works out which regulatory obligations apply to your product and hands your team traceable engineering tasks, with an approval gate at every step.

Inside the product

The whole passage, in one glance.

01 · Context

Likely to be accessed by under-18s?

Yes — includes minors

→ assumption recorded

02 · Obligation

UK-AADC-STD-11

Profiling switched off by default

Applicable · rationale attached

03 · Ticket

COMP-52backlog · ready

Default profiling off for minors

Profiling disabled before first session

Opt-in requires verified guardian

04 · Export

audit-bundle.json

sha256 · 8f3a92c1…e47b19

Sealed ✓ tamper-evident
1

Step 1

Context intake

Describe your product

Answer questions about your service: what it does, who it affects, what data it processes, whether it deploys an AI system. Your answers become structured assumptions. Applicability mapping is rule-based and deterministic; no model decides what applies to you.

  • Do you deploy an AI system in a professional capacity?
  • Which Annex III area? Are you a public body?
  • Does your product target or is it likely to be accessed by users under 18?
  • What personal data do you process, and on what legal basis?
Output: Categorised assumptions: AI role, Annex III area, body type, audience, data, jurisdiction
2

Step 2

Obligation mapping

Rules decide applicability. You confirm.

Each obligation carries trigger conditions that are evaluated against your answers, deterministically: the same answers always produce the same mapping, and no model decides what applies to you. AI is used only for the interpretation text and an optional narrative, each labelled as such. You confirm, override, or flag each mapping, and every decision requires a rationale.

  • Obligation ID linked to regulation source
  • Applicability status with reviewer badge
  • Plain-language rationale for every decision
Output: Reviewed and approved obligation mappings
Approval gate: obligation mappings
3

Step 3

Engineering tickets

Obligations become work items

Each applicable obligation produces prioritised work items with acceptance criteria, priority levels, and full traceability back to the source regulation.

  • MoSCoW priority assignment (Must / Should / Could)
  • Acceptance criteria for each item
  • Multiple item types: Feature, Edge Case, QA Test, Logging
Output: Sprint-ready backlog items
Approval gate: work items
4

Step 4

Export and audit

Ship with confidence

Export to Jira, Linear, or as audit-ready packages. Every export includes cryptographic integrity verification and complete decision trails.

  • Push directly to Jira or Linear
  • SHA-256 integrity verification, checked in CI with the @reg-to-backlog/verify CLI
  • Self-contained HTML audit packages, OSCAL JSON and the Dutch Algoritmeregister entry
Output: Audit-ready compliance documentation

Built on three principles

Human in the loop

Rules map, humans decide. Applicability is deterministic, and every mapping still requires explicit approval with a documented rationale. Where AI writes interpretation text, it is labelled as AI-written.

Ambiguity detection

When regulatory text is unclear, Landfall flags it for human resolution. No silent interpretations.

Full traceability

Every work item links to an obligation, which links to specific regulation text. Complete audit trail.

The harbour

Ready to see it
in action?

Create your first project in minutes. Landfall reads the regulation; your team builds the product.