Skip to content
RC.00Deployment recommendation engine

Tier + substrate + modules, tailored to your programme.

Seven inputs in. A 7-section streaming recommendation out — recommended tier (L.01-L.06), substrate (DM.01-04), module bundle, engagement cadence, indicative pricing pointer, vertical-specific caveats and explicit next-step routing. Downloadable as Markdown; logged to the AI artefact registry.

RC.SFLive recommendation engine
No recommendation yet · fill inputs and generate

Tier + substrate + module-bundle recommendation tailored to your programme inputs. Cross-references the engagement cadence, indicative pricing, and the next-step routing through sandbox / artefact-generator / consultation.

RC.OT · streamed via /api/ai/ask

RC.BTFour guarantees

Recommendation, not a suggestion engine.

  • RC.B1

    Tailored to your inputs

    Seven inputs drive the model — programme, jurisdiction, vertical, scale, sensitivity, modules-of-interest, organisation. Output is structured against the FlyttGo tier ladder, deployment substrate menu and module catalog.

  • RC.B2

    Anchored to real shapes

    Recommendations cite tier codes (L.01-L.06), substrate codes (DM.01-04), engagement cadence (SE.D1-3) and section codes that route to the right page. Reviewers can trace every claim.

  • RC.B3

    Provenance-tracked

    Every recommendation logged to /governance/ai/artefacts with a SHA-256 content hash. Reviewers verify the artefact has not been tampered with; ties into the broader AIBOM (AG.00).

  • RC.B4

    Routes onward, not in circles

    Output ends with explicit next-step routing — sandbox / artefact-generator / consultation — so the recommendation kicks off real engagement rather than leaving the visitor at a download.

RC.NXWhere the recommendation goes next

A recommendation is the start of a real procurement engagement.

The output is structured to route onward. The four pathways below take a recommendation into the surfaces it cites.