SOSX Civic

SOSX Civic is coming soon. It will help councils, combined authorities, public bodies, charities and policy teams see a place, or a policy area, as one system: the services, the people they serve, the partners, the money and the conditions around them, and the influences that join them up. On that one map you will be able to test the options side by side, with every figure taken from a cited source and every perspective given its strongest case. SOSX Civic will lay out the options, the evidence and the trade-offs. People decide.

None of what follows needs us. Public bodies already have good method for problems that cross services and organisations: the Government Office for Science's systems thinking toolkit for mapping a problem before solving it; soft systems methodology, which records how each group of stakeholders sees the situation; boundary critique, which asks who and what has been left outside the frame; HM Treasury's Green Book for appraising options, with business as usual always on the list; the Magenta Book for planning an evaluation; and a theory of change linking what is done to what it should achieve.

What makes them hard is the time each one takes, and that they live in different documents, built by different teams, which go out of date the moment an assumption moves. SOSX Civic will bring them onto one map that stays current.

• One map of the place or the policy. Services, population groups, partners, resources, conditions and outcomes as factors, joined by influences and loops, drawn as a place map or a policy map.
• Every figure traceable. Each number from a cited, published source or a deterministic calculation, with the strength of its evidence shown beside it. Lived experience is recorded as evidence in its own right, not as a lower rung.
• Every view, with its strongest case. Each stakeholder group's view recorded and attributed, and its strongest case stated before any options narrative is written.
• A boundary review. Who is inside the map, who is outside it, and whose voice is missing. What ought to be in scope is answered by people.
• Options compared like for like. Business as usual and do-minimum always included, appraisal aligned with HM Treasury's Green Book, sensitivity tests and switching values, and no single composite score.
• Demand and the budget, projected. Service demand projections and a budget-gap scenario under each option.
• Through to evaluation. A theory of change and an evaluation plan for the chosen option, with equality questions and data gaps raised for a person to conclude.
• Balanced packs for scrutiny. Every perspective with its strongest case, the evidence, the uncertainties and a statement of how AI was used, with a decision trail for scrutiny and freedom of information requests.
• Built for public-sector rules. Aggregate data only, with small numbers suppressed; a pre-election period switch that holds publication; a full text alternative to every map; and a record ready for the Algorithmic Transparency Recording Standard.

Public decisions are contested, and the analysis behind them has to be fair and be seen to be fair. SOSX Civic will be strictly non-partisan. It will set out options, perspectives and trade-offs, and it will not tell anyone which option to choose. Every AI-written passage will be checked for balance and for loaded wording before it can be published, and a person approves each item.

It will be built for public policy analysis, public-service design, option appraisal and public deliberation. It will refuse requests for election material, for profiling people, and for decisions about named individuals. More than fifty specialised AI agents will do the research and the drafting; officers, members and boards make the decisions.

Harrowmere is invented, to show the idea. It is not a real council and none of its details describe a real place.

In Harrowmere, the number of children looked after has risen for three years, and external residential placements are driving an overspend in children's services. SOSX Civic will map the place: children and families, foster carers, social workers, early help, the provider market, health partners and the inspectorate, and the loops between them. One is easy to recognise: too few in-house foster carers means more external placements, higher costs and less money for the early help that keeps families together.

Each group's view will be recorded with its strongest case, and the boundary review will ask whose voice is missing, such as care leavers. Five options will be compared side by side: carrying on as now, renegotiating the provider framework, more in-house fostering, a regional cooperative with neighbouring councils, and investment in family help. The scrutiny committee will get a balanced pack with every source cited. The council decides.

• Councils: strategy, transformation and finance teams, and children's services, adult social care and public health.
• Combined authorities and their analysts, and place partnerships in integrated care systems.
• Central government policy and analytical teams.
• Think tanks, charities and NGOs, and parliamentary research services.
• Democratic services and communications officers, who need balanced evidence and a clear record of how it was reached.