AI in Action

Five New Zealand business case studies.

From high-volume credit analysis to workshop diagnostics. Real data. Real implementations. Measurable impact.

Case Study 01

AI Credit Decision Pipeline

Equipment leasing & finance (subprime, high-volume applications)

× The Problem

A single credit manager manually investigated every lease application — each taking 20 to 120 minutes. On busy days, five or six applications stacked up while sales teams waited on bond confirmations, creating a bottleneck that capped growth.

What We Delivered

A 4-stage AI credit analysis pipeline (Identity → Financial → Cross-Document → Verdict) that maps 16 discrete checks directly to the credit manager's existing procedure. The system ingests credit reports, bank statements, and driving records, then produces a scored recommendation with full audit trail — grounded against source documents so every finding is traceable.

The Result

Assessment time dropped from hours to minutes. The business can now process ten times the application volume without adding headcount. The CFO described the initial demo as having "exceeded my expectations."

Case Study 02

AI Fleet Diagnostics

Transport refrigeration servicing (nationwide fleet, legacy workshop systems)

× The Problem

Technicians spent 20+ minutes identifying parts and searching PDF manuals, often with grease-covered hands. Service history was trapped across disconnected legacy systems (Azure SQL, invoicing, technician portal), making it near-impossible to spot patterns like repeat faults or declining unit health.

What We Delivered

An AI diagnostics layer built on top of the existing 69-table legacy database (199,000 jobs, 28,000 units). The system provides instant unit intelligence, since-last-job analysis, and trend detection — classifying every insight as data-backed, calculated, or AI-estimated so workshop staff can trust what they see.

The Result

A working demo on real production data was delivered within one day. The client moved from first conversation to paid engagement in three days. The system is live and in active use.

Case Study 03

AI Opportunity Roadmap

Multi-discipline engineering (130+ staff, 5 offices, 9 technical disciplines)

× The Problem

Rapid growth was driving data fragmentation and administrative burden across planning, structural, civil, surveying, geotechnical, environmental, traffic, fire engineering, and draughting teams. Leadership knew AI could help but had no structured view of where to start or what the realistic return would be.

What We Delivered

A comprehensive AI opportunity analysis: five in-depth research documents covering company profile, operational process inventory, workflow lifecycle analysis, and a prioritised AI enablement roadmap. Delivered alongside interactive data visualisations (Operational Atlas) mapping every process across all nine disciplines to specific AI capabilities.

The Result

The analysis identified 8,550 hours per year of automatable work worth an estimated $1.54M annually across the business — giving leadership a clear, evidence-based roadmap for phased AI adoption.

Case Study 04

AI Project Health Analysis

Construction technology (SaaS platform aggregating data from Procore, Jobpac, MS Project, HammerTech)

× The Problem

The platform had rich project data across a 319-table database but no AI-powered analysis layer. Leadership wanted to surface margin erosion risks and project health insights automatically — not just dashboards, but intelligence that explains *why* a project is trending off-track.

What We Delivered

An AI analysis pipeline that ingests real project data (contracts, costs, forecasts, variations) and returns structured health assessments with field-level lineage — every number traced back to its source table and column. Built on top of the client's existing Azure infrastructure with no changes to their production code.

The Result

Analysis runs in under 60 seconds on live project data. Phase 2 is now signed to embed the AI layer directly into the SaaS product for customer-facing use.

Case Study 05

AI Content Factory

Insurance brokerage (personal, commercial, fire & general)

× The Problem

The brokerage had 30+ brand images ready but no capacity to turn them into platform-specific social content at scale. Creating, captioning, scheduling, and publishing posts across Facebook, Instagram, and LinkedIn was manual, slow, and inconsistent.

What We Delivered

An AI-powered content factory that generates platform-optimised captions, produces AI imagery aligned to brand guidelines, and publishes directly to all three social platforms. The system includes a 14-day schedule view, mobile-friendly UX, and synced post management (including deletion) across channels.

The Result

The brokerage moved from zero consistent social presence to automated, multi-platform publishing — turning existing brand assets into a steady stream of professional content without hiring a marketing team.

Ready to build your own case study?

Start with a clear prototype on your actual business data in just 2 weeks. Eligible for up to 50% government co-funding via the RBPN AI Advisory Pilot.