Regulatory compliance engines for public sector.

Policy-as-code for government. Decisions that can be explained, evidenced, and reviewed — by Parliament, NAO, or the people affected.

Request a private briefing→All regulatory work→

Overview

Regulatory compliance engines infrastructure for public sector, built to the standard institutions in this sector are required to operate.

XVICA designs, builds, and operates this layer for public sector clients in the UK, US, Canada, and Australia. The work is specified against the regulatory regime, the operational profile, and the examination expectations of this sector before any code is written.

01Why it matters

What public sector cannot get wrong here.

Every policy decision must be explainable, in public, years later.
Legislative changes take effect on a specific date and must be implemented by it.
FOI requests can ask why a specific decision was made for a specific person.
A wrong decision published in the press can force a Ministerial statement.

02Regulatory posture

Named regimes, mapped controls

Regulatory requirements are translated into explicit control requirements, then mapped to tests and evidence collection. Nothing is implied.

Decision-making standards

Public law principles, the Algorithmic Transparency Recording Standard, the Ministerial Code, and judicial-review-informed design.

Data & privacy

UK GDPR, Data Protection Act 2018, DPIAs, and the ICO's framework on automated decision-making.

Accountability

NAO value-for-money framework, Managing Public Money, and departmental Accounting Officer obligations.

03Reference architecture

Design decisions distinctive to this intersection

Components and design choices that recur across our work for this sector. Each deployment is specified individually.

Policy-as-code

The policy is the rule set. Changes to policy are changes to versioned code, with author, reviewer, and effective date.

Reason-giving at the decision

Every decision produces a human-readable explanation alongside the outcome, using the language of the policy rather than the engine.

Appeal and correction pathways

Decisions can be reopened; corrections are recorded against the original, not as overwrites.

Algorithmic transparency

Where models inform decisions, their role is disclosed in the form the ATRS requires.

Historical reproducibility

A decision made in 2026 can be reproduced in 2031 using the rules and data as they stood in 2026.

04XVICA's approach

How we work in public sector.

Public-sector decision-making has a property commercial systems rarely have: the same decision may be litigated, reviewed, reported on, and re-examined years after it was made, sometimes with the people who made it no longer in post. We build with that horizon in mind. Policy is authored as versioned rule sets with named reviewers and effective dates; decisions carry their reason-giving in the language of the policy rather than the engine; corrections are additive rather than destructive. The practical consequence is that Judicial Review, FOI, and NAO inquiries each become queries against the platform rather than reconstruction projects, and the department retains the institutional memory of why a decision was made even as the people and the administration change around it.

05Engagement