Systems that stay accountable to the people responsible for them.

Post Trade Platform

Replacing a low-code SaaS with an in-house build, while strangling the API beside it into services.

Governed agent access

An MCP layer giving AI agents authorised, scoped and audited access to data they must never reach directly.

Quorum

A governed memory layer so agents remember what your team already decided.

A proof of concept, still defining where the value is. Today, tracing what someone did across separate platforms, systems and channels is a manual job that takes around 30 days. This unifies that trace into one place, so activity carried out with ill intent can be detected and reported in under five minutes. An LLM first looks for an innocent explanation for each event, then reviews the whole series to judge whether it looks like a one-off or a deliberate pattern. A supervisor makes the call.

Replaced a low-code SaaS post-trade platform with an in-house build, while strangling the internal API that grew up beside it into services.

A governed memory layer for AI agents. Stores engineering decisions and business requirements with authorship, rationale, conflict detection, authority weighting and a signed audit chain.

One surface over HR systems, building services, approvals and a SharePoint CMS — with consistency and accessibility enforced at a boundary rather than requested.