Case Studies

Real Results, Real Impact

Explore how we've helped organisations across industries harness AI to solve complex challenges and deliver measurable outcomes.

Case studies

Case studies and delivery work

These write-ups cover client engagements and first-party delivery work. Each one explains the operating constraint, the system design and the accountable delivery approach.

Case study

Governed agent tooling: one platform, single toolset

An assistant that can call an enterprise platform's management API can build and change production systems. Nothing in that API decides whether it should. This is the layer that does.

AI agentsMCPTool designAuthorisationMulti-tenancy

Case study

Automated regression harness for retrieval-grounded agents

A change to a prompt, a retriever or a base model can quietly break an answer that used to be correct. We stopped checking that by hand.

EvaluationRetrievalCI/CDInternal tooling

Case study

Layered guardrails across voice, web chat and messaging

A policy that holds on web chat and quietly softens on voice is not a policy. The layers, the order they run in, and what the assistant does when one of them fires.

AI agentsGuardrailsMulti-channelVoiceTesting

Case studyIndustry: Technology

Testing LLM systems before release

Delivery work for testing retrieval-grounded LLM systems before release, with emphasis on hallucination control, context preservation and multilingual terminology.

LLM testingEvaluationRelease governanceRetrieval

Case studyIndustry: Technology

A shared quality model for enterprise AI

Delivery work for unifying AI quality assurance across customer service, document processing and decision-support workflows.

AI testingQuality assuranceRegressionGovernance

Case studyIndustry: Insurance

Answering insurance enquiries, escalating everything else

The design settled what would never be automated before it settled what would. Everything after that was a question of evidence: what the answer was built from, and whether it could be shown.

Contact centreGeneral insuranceRetrieval groundingEscalation designAI governance

Case studyIndustry: Financial Services

Automating routine forms, designing escalation that matters

A pension fund processed member forms by hand. Conversational AI and RPA took the routine work, and manual verification was an element of the design, not a fallback from it.

Pension FundProcess AutomationConversational AIRobotic Process Automation

Case studyIndustry: Healthcare

Triage support that gathers and escalates; a clinician decides

The design question was never how much of triage could be automated. It was which part of it must never be, and how the system behaves at the edge of what it is allowed to do.

Health servicesClinical governanceEscalation designHuman accountabilityTriage support

Case studyIndustry: Financial Services

Fraud detection with review controls

Delivery work for real-time fraud detection, false-positive control and review-team prioritisation in financial-services workflows.

Fraud detectionFinancial servicesRisk scoringReview workflow

Case studyIndustry: Retail & E-Commerce

Personalisation with merchandising control

Delivery work for recommendation quality, merchandising control and customer-experience measurement in retail journeys.

PersonalisationRetailRecommendationsJourney measurement

Case studyIndustry: Retail & E-Commerce

Inventory decisions with exception review

Delivery work for demand forecasting, replenishment support and exception review across distributed retail inventory.

Inventory intelligenceForecastingRetail operationsException workflow

Case studyIndustry: Manufacturing

Predictive maintenance with planner review

Delivery work for equipment-health monitoring, maintenance prioritisation and explainable intervention planning.

Predictive maintenanceManufacturingAsset healthDecision support

Case studyIndustry: Insurance

The journey review came before the automation

Five problems were named across voice, web and email, and one of them was the process rather than a system. The review of the customer journey ran before any element was chosen, and the elements that followed had to run on the technology and the training already in place.

Design ThinkingService DesignConversational AIMachine LearningDocument ExtractionRobotic Process AutomationMicrosoft Dynamics & 365Insurance

Real Results, Real Impact

Case studies and delivery work

Governed agent tooling: one platform, single toolset

Automated regression harness for retrieval-grounded agents

Layered guardrails across voice, web chat and messaging

Testing LLM systems before release

A shared quality model for enterprise AI

Answering insurance enquiries, escalating everything else

Automating routine forms, designing escalation that matters

Triage support that gathers and escalates; a clinician decides

Fraud detection with review controls

Personalisation with merchandising control

Inventory decisions with exception review

Predictive maintenance with planner review

The journey review came before the automation

Ready to shipEnterprise AI?