SOSX®, AI-enabled systems thinking, for any sector.
For public-sector digital and assurance leads
Public-Sector Digital / Assurance Lead
SOSX® is AI-enabled systems thinking, the definition, research and analysis of systems-thinking problems, in any sector. Built initially for the concept design phase of MBSE and for Innovate UK's Agentic AI Pioneers competition, today SOSX is a sector- and subject-agnostic agentic AI platform for systems-thinking research, analysis and optimisation. Its specialised AI agents map a problem as a connected network, research it from cited sources and run grounded, auditable analyses, on a platform built against ISO 42001, ISO 27001, the EU AI Act and NIST AI RMF, producing the governance evidence those frameworks ask for: agentic AI that is safe for enterprise.
The discipline: what the assurance standard already asks
You hold yourself to this already; we are stating it so you can see what we are proposing to help with. Options appraised against a stated counterfactual, so "do nothing" is costed rather than assumed away. Assumptions declared with their sources, and sensitivities shown where the answer turns on them. A decision trail that records who advised what, when and on what evidence. Public money treated as public: the evidence has to survive being read by someone who was not in the room and does not share your priors. Proportionate scrutiny of suppliers, including their claims about their own technology. And a record kept for long enough that a successor, an auditor or a committee can reconstruct the reasoning years later.
None of that requires a platform, and the parts of it that go wrong are rarely the analysis. They are capacity and currency: nobody has time to map the estate, and the map goes out of date faster than it can be rebuilt. The expensive part is not drawing the model; it is building it and keeping it true.
Scepticism is the job, so we start there
If you lead digital, procurement or assurance in a government body or regulator, you must both adopt AI responsibly and evidence that adoption to oversight bodies, with limited internal capacity to map a complex system estate, and a well-earned scepticism of AI-vendor claims. That scepticism is not an obstacle to this proposition; it is where the proposition starts.
SOSX is AI-enabled systems thinking that shows its working. Every claim in an analysis cites a catalogued source. Judgement, formal proof and computation are kept visibly separate, so a reviewer always knows what kind of statement they are reading. Every AI interaction is audit-logged. The stages that decide the model pass a human approval gate, and an automatic build is held by review gates it cannot clear on its own. For an estate no internal team has the capacity to map, SOSX builds the connected model and keeps it analysable, and the outputs are hand-over-able artefacts: evidence packs whose language maps to the frameworks oversight bodies recognise, from a platform built against those same frameworks.
We also apply to our own claims the discipline you apply to vendors': SOSX aligns with and produces evidence for those standards: we never say "certified to", because that is not ours to say.
The markers procurement actually checks
Digital Tactics, the company behind SOSX, is a 15-year UK SME (founded 2011, Hove) with a listing on the government's digital procurement frameworks, Cyber Essentials certification and AWS partner status: more than 170 client projects, nine products. The platform's referenceable engagement is live enterprise MBSE trials with a large global aerospace manufacturer. The framework-by-framework detail is on the compliance page; the company story is on the about page.
The question the oversight body will ask
When the oversight body asks how this analysis was produced, what do I hand them? With SOSX the answer is an artefact, not an assurance: the sources, the approval trail, the audit log, produced in hours rather than weeks of assembly.
Go deeper
This sector has no companion book of its own: the central book is where the argument it rests on is made in full.

Bring us a real assurance question
Bring us a real assurance question (the estate you don't have capacity to map, the AI adoption you have to evidence) and we'll show you exactly what the hand-over artefacts look like, caveats included. Everything SOSX asserts arrives with its sources attached, so nothing we show you has to be taken on trust.
