AI case studies

See how practical AI becomes real business improvement.

Explore AI case studies showing how Changeable turns operational challenges into working systems, clearer decisions and measurable outcomes across AI strategy, software, automation and process improvement.

How to read these examples

Not generic AI showcases. Business problems shaped into practical systems.

The work shown here is not about chasing tools for their own sake. Each example starts with the business context, the operational constraint, the governance risk or the decision that needed better support. Where projects involve personal information or public-sector use, delivery is informed by guidance from the Office of the Privacy Commissioner and New Zealand Government digital guidance.

Some projects focus on AI strategy and readiness. Others involve AI governance, knowledge systems, workflow automation, AI agents, data models or AI-enabled software development.

Clarity before build

Projects are grounded in the business problem, process, risk and expected value before solution design begins.

Human review built in

AI use is designed with accountability, source traceability, review points and practical governance controls.

Systems that fit the work

The solution is shaped around existing workflows, team capacity, user behaviour and operational constraints.

Measured improvement

The aim is useful change, including faster onboarding, less manual effort, better consistency or more confident decisions.

Selected work

Selected AI case studies and delivery work

Examples across AI knowledge systems, governance, behavioural simulation, document intelligence, secure extraction workflows, forecasting agents and workflow automation.

AI Operating AssistantAI Agents

Briefable: A Connected AI Operating System

A modular AI operating assistant connecting email, calendar, files, contacts and pipeline into one governed daily briefing, with human review kept visible before anything leaves the platform.

View project
One daily brief drawn from five connected sources
Security audited before any external tester access
Human approval required before any external action
Workflow AutomationCompliance

Purpose-Driven Workflow Automation

Standardising operational intake and funding compliance reporting for a regional New Zealand social impact enterprise through comprehensive stakeholder elicitation and targeted process design.

View project
18 hours of weekly administrative relief per coordinator
100% contract compliance alignment
Xero-connected validation routing
Decision IntelligenceKaupapa Māori

Pūtake Labs: Evidence-Disciplined Decision Intelligence

A bicultural decision intelligence practice built on eight integrated analytical Labs, a strict evidence-rating discipline and a living-analysis engine that tracks findings over time.

View project
Eight Labs across Understand, Test and Implement stages
Every finding rated High, Moderate or Low, no exceptions
Method stress-tested against hard material before software was built
Capture-to-ComplianceReal Estate

Snapsure: AI-Empowered Real Estate App

A capture-first AI platform that turns photos and voice notes from a property walkthrough into compliant inspection reports and marketing-ready listing content from the same capture engine.

View project
One capture engine powering inspections and listings
Purpose-built Healthy Homes compliance assessment
Tested against 58 defined user stories before release
Document ExtractionXero

Commercial Lease Audit Automation

Secure AI-assisted lease audit automation for extracting commercial property terms, protecting margins and supporting governed financial workflows.

View project
Structured extraction of lease terms, CPI clauses and renewals
Human-reviewed workflows with secure local data handling
Xero-connected alerts to reduce missed escalation
Contract IntelligenceGoverned AI

Building ObliTracker

A governed AI contract intelligence engine designed to extract obligations, key dates, risks and client-ready deliverables from complex agreements.

View project
Contract obligations, dates and risk points extracted
Governed AI workflow with human review and traceability
Client-ready summaries, registers and deliverables
Knowledge RetrievalRAG

Retail Knowledge Intelligence System

AI-powered retail knowledge system using RAG and Gemini to improve training, compliance and consistency across a multi-site retail environment.

View project
Centralised retail knowledge using RAG and Gemini AI
Reduced onboarding time across NZ stores
Role-based access with auditable delivery
AI GovernanceEthics

Designing and Implementing Responsible AI Governance

Enabled responsible AI adoption by aligning governance, ethics, business-unit behaviours and day-to-day AI use at scale.

View project
Governance-aligned AI use cases and decision controls
Confidence gaps and adoption barriers mapped
Human-in-the-loop workflows and safeguards
Behavioural AISimulation

Beyond Feasibility: Market Validation with Behavioural AI

Changeable built a behavioural AI capability to test human and AI interaction dynamics in controlled, data-rich conditions.

View project
Persona-based behavioural drivers and friction points
Trust factors, assurance and perceived risk modelled
Micro and macro forces shaping feasibility
Document IntelligenceAutomation

AI-powered Document Intelligence

Document intelligence and automation to extract information, categorise material, flag risk and route work more consistently.

View project
Extract key data points from incoming documents
Scan and categorise documents for operational use
Flag compliance risks and route approval tasks
AI ForecastingAI Agents

Enterprise AI Forecasting and Strategy Agents

AI forecasting and strategy agents designed to support faster decision-making, predictive insight and structured strategic analysis.

View project
Data integration across operational inputs
Predictive modelling and scenario analysis
AI agents operating as virtual strategy support
Workflow AutomationSME

SME Workflow Automation

Workflow automation for small teams that need to reduce manual work, speed up core processes and improve customer experience.

View project
30 to 40 percent reduction in admin tasks
20 percent decrease in operational costs
Faster invoicing and improved customer experience
Process ImprovementBA

Leanable: Design and Build of an AI-Led Process Improvement Platform

An AI-supported process improvement platform shaped around business analysis, workflow mapping, stakeholder input and practical improvement planning.

View project
Maps current workflows and identifies process friction
Supports practical future-state process design
Turns improvement work into clearer deliverables
Public InterestAnalytical Engine

Observed: An Evidence-Based Organisational Research Platform

A methodology-governed analytical engine for public-interest research, evidence review, accountability analysis and human-reviewed outputs.

View project
Structured public evidence review and source handling
Governed analytical workflows with human oversight
Defensible outputs for accountability research

What these projects show

Changeable’s work sits across the practical middle ground where business analysis, process improvement, AI implementation and governance need to connect.

Knowledge systems

Turning scattered organisational knowledge into trusted, searchable and source-backed support for real teams.

Governance and safety

Helping organisations adopt AI with clear risk thresholds, human oversight and practical decision controls.

Simulation and validation

Testing behaviour, trust, feasibility and adoption dynamics before leaders commit to higher-risk decisions.

Automation and delivery

Designing workflows, document systems, AI agents and operating models that reduce friction and improve consistency.

Start with a free discovery session.

A no-obligation conversation where we listen to where you are, what you are trying to achieve and what is getting in the way. Get in touch to start.