Skip to main content

References

Some of our work we can show you. The rest we walk through on a call.

These clients run in regulated or commercially sensitive settings, so they appear without names. The situation, the build, and the figures come straight from the delivery. Name the capability you are weighing and we will match it to comparable work.

Selected delivery, anonymised by agreement

Each client below works in a regulated or commercially sensitive sector and is shown without a name, company size, or region. Everything else, the starting point, the work, and the numbers, comes from the engagement itself.

Manufacturing · Chemical R&D

Generative AI knowledge platform · RAG on Azure AI Search

The situation

A global chemical company held its R&D knowledge across separate sites and systems. Researchers could not get answers on chemical processes, material properties, or technologies without searching by hand, and laboratory equipment data sat unconnected across locations.

What we did

We designed and built a generative AI platform for more than 2,000 R&D users. It runs Retrieval-Augmented Generation on Azure AI Search, with a modular backend API orchestrating specialised agents for knowledge retrieval, document analysis, and context enrichment.

Result

  • One generative AI platform serving 2,000+ R&D users across globally distributed sites.
  • Natural-language querying of chemical processes, material properties, and technologies that previously needed manual search.
  • Laboratory equipment classified automatically and enriched with metadata (manuals, specifications, location), making cross-site resource discovery possible for the first time.
Hospitality

Cloud analytics platform · Microsoft Fabric and Power BI

The situation

A hotel group had no single view of occupancy, revenue, and forecasts. The data sat in separate property-management systems, so the business had no current picture to plan against.

What we did

We designed and built a cloud analytics platform on Microsoft Fabric. It ingests data automatically from multiple hotel PMS systems over REST APIs, models a unified KPI set (ADR, RevPAR, pick-up, forecasts), and presents interactive Power BI dashboards with row-level, role-based access.

Result

  • Platform live within 12 weeks.
  • In use by more than 100 people.
  • One current view of occupancy, revenue, and forecasts across the group, replacing fragmented per-system reporting.
Manufacturing · Chemical

AI computer vision · Power Platform and IBM MAXIMO

The situation

On the shopfloor, finding the right spare part was slow and error-prone. Long search times and wrong orders drove avoidable downtime costs in production.

What we did

We built a mobile app for spare-part identification at the point of work: smartphone image capture with AI classification, a domain-specific computer-vision model trained on more than 4,000 parts, and integration with IBM MAXIMO for inventory, location, and pricing. Each prediction surfaces with a confidence score and alternatives.

Result

  • Computer-vision model trained on 4,000+ parts.
  • Spare-part identification moved from manual search to on-device image capture.
  • Lower downtime costs in production from fewer wrong orders.

Most of our work cannot be shown at all. Energy infrastructure, financial services, and healthcare clients sign us under NDA, and that holds after the work ships.

Name the capability you are weighing, an Azure architecture review, a Power BI rebuild, a first AI use case, and we will walk you through a comparable engagement: where the client started, what we changed, and the number that moved.

Name the proof you need.

A short call with the architects who did the work. We will match it to a reference you can weigh.