Governance-first AI, in the real world.

The challenge

Useful market intelligence is buried across dozens of fragmented sources, and most monitoring tools drown teams in hype rather than surfacing the few signals that actually matter for a decision.

What we did

We built an automated radar that ingests dozens of public sources, including news, hiring, company filings, events and regulatory feeds, then applies a weighted signal framework to score organisations on genuine intent rather than noise.

What we built

A two stage language model pipeline does the work. A low cost pass extracts and tags every item, then a higher quality pass writes concise, decision ready briefs. Entity resolution merges duplicates and public company data confirms size and fit. It runs weekly on public or permissioned data only, with scoring that stays explainable.

The challenge

A popular internal knowledge assistant was trapped behind per seat vendor licensing, so every new user meant another license. [TODO: confirm the exact pain and who felt it.]

What we did

We designed a license independent architecture inside the client's own cloud tenant: Azure AI Search plus Azure OpenAI, documents never leaving their environment. [TODO: confirm final architecture variant.]

What we built

[TODO: what shipped, where it runs, who owns it, what it replaced. Note migration status: proposal / in build / live.]

The challenge

Teams wanted on demand answers from brand positioning, market research and campaign results, but the tool that did it depended on per seat licensing. [TODO: confirm.]

What we did

Same governance first pattern as CORA: retrieval over the client's research corpus, rebuilt to run inside their tenant without per seat licenses. [TODO: confirm.]

What we built

[TODO: what shipped and how it is used. e.g. a mandated step in NPD and campaign work.]

The challenge

A global dry bulk operator with complex, distributed operations wanted to know where AI could reduce manual information work without disrupting workflows where a wrong answer carries real operational risk. Enthusiasm existed across teams, but there was no shared, evidence based view of where to start.

What we did

We ran a structured Discovery across operations directors, team leads, commercial and technical functions, combining in depth interviews with role tailored questionnaires. We mapped current state, surfaced pain points, and scored candidate use cases on value, feasibility and risk.

What it produced

Discovery produced a prioritised use case roadmap and a technical report leadership could act on. The recommended first build is a governed Knowledge Assistant, a retrieval system over the company's own operational knowledge, designed for Azure with human oversight and explicit ownership.

The challenge

A global brewer had real momentum with internal AI assistants, but adoption was constrained. The most popular tools depended on per seat vendor licensing, and teams across the business wanted a clear, governed way to scale what was already working.

What we did

Discovery mapped where AI could create value across functions. We then designed an architecture that runs inside the client's own cloud tenant, so their documents never leave their environment, and the capability no longer depends on per seat licensing.

What it produced

AI now supports Marketing, Corporate Communications, M&A and regional operations. The internal assistants (see CORA and FLOID under Builds) were re engineered for ownership: client owned data, client owned tenant, governance built in from day one.

The challenge

A facilities and property management group ran complex cost allocation, maintenance and statutory compliance work through spreadsheets and disparate tools. They wanted to know where AI could cut manual effort without disrupting the trusted open book model their tenant relationships depend on. Part of the team did not work in English.

What we did

We ran a compressed two week Discovery with role tailored questionnaires and interviews delivered in the client's own language. We mapped the cost allocation engine, the claims and warranty pipeline, statutory compliance tracking and tenant reporting.

What it produced

Discovery delivered a technical report plus an AI capability catalogue, a menu of scoped, costed build options matched to the workflows that lose the most time, each with its own value and feasibility read.