Square One is an enterprise platform for organisations that run on Microsoft 365. It gives AI, documents, operational data and organisational knowledge one governed place to live: permission-aware, evidence-grounded and auditable. The premise is that most work inside an organisation is ungoverned, and that AI has made that an urgent problem rather than a new one.
Inhouse CX shaped the product direction and built the platform, from the data model and architecture through to a live, deployed system. The work spans a Go backend, a React front end, a Microsoft 365 integration layer, a mobile client, and the public marketing site.
The thesis
The starting point is a plain observation: staff across an organisation already use AI, usually in personal accounts, because it helps them work. What is missing is a governed place for that work to happen. Without one, organisations lose the ability to say what evidence an output rests on, who approved it, or whether it can be trusted.
Square One treats this as a governance problem, not a chatbot problem. Its position is that automation should handle deterministic work, AI should be used only where reasoning adds value, and everything should stay permission-aware, evidence-grounded and on the record.
What Inhouse CX built
Inhouse CX shaped the product and built the platform across its major parts:
- A Go backend (modular monolith) on PostgreSQL, with vector search for retrieval.
- A React and TypeScript web app.
- Native iOS and Android apps.
- A Microsoft 365 integration layer over Microsoft Graph (SharePoint, Exchange), with Entra ID identity and permission-aware retrieval.
- Evidence-grounded document generation, a governed assistant, and operational modules (asset, hire and reconciliation workflows).
- A marketing site with structured-data-led SEO.
- Cloud-native deployment on Azure.
Approach
A few principles run through the build. Governance is inherited from Microsoft 365 rather than reinvented, so the platform strengthens an organisation's existing permissions instead of working around them. AI is model-agnostic and sits behind a policy gate, so providers can change without changing the operating model. Business artefacts stay traceable to their evidence, approval-gated, and auditable after the fact.
Inhouse CX is a Microsoft Solutions Partner in Data & AI and Digital & App Innovation on Azure.
Where to next?
This work draws on our Systems & Platform Development capability.