Client
Large public sector organization
Industry
Public sector
Primary goal
Accelerate end‑to‑end solution validation during core platform modernization
Technologies
Microsoft Azure, GPT-4o, Playwright, GitHub, SSIS, Cosmos DB, JIRA, APIs
Make transformation faster
Large‑scale digital transformation requires significant time, effort, and specialized expertise. Core activities such as data migration, testing, and validation are often manual, resource‑intensive, and sequential. This heavy lift can slow momentum and, in some cases, make transformation harder to realize.
AI creates an opportunity to rethink how transformation is executed. Rather than applying AI only to the future‑state solution, organizations can embed AI directly into transformation operations. Activities that once required extensive manual effort, such as data mapping, extraction logic design, and test case development, can now be augmented and accelerated using AI.
Here’s how KPMG worked with a large public sector organization to embed AI directly into transformation execution, accelerating end‑to‑end testing while maintaining appropriate governance and human oversight.
50%
reduction in test scenario drafting effort
75%
reduction in Playwright scripting effort
40x
faster test execution (from ~40 minutes to ~1 minute)
85%
overall coverage versus user‑written baselines
Client transformation journey
- Before
- During
- After
Manual, capacity‑constrained testing during platform modernization
At the core of the program was a complex source‑to‑target transformation: Migrating and converting legacy data into a modern core platform environment. Ensuring that legacy data was accurately mapped, transformed, loaded, and functioning correctly in the target system required extensive validation.
Testing extended beyond standard release cycles. It required validating source‑to‑target data mappings and transformation logic, converted data integrity within the target environment, end‑to‑end business process functionality using converted data, and ongoing regression testing as configurations and releases evolved.
Because data conversion quality could only be proven through downstream solution behaviour, testing became the primary control point for validating legacy transformation at scale.
Traditionally, business analysts, QA analysts, developers, and data specialists drafted detailed test scenarios, built scripts, executed manual tests, reconciled defects, and maintained traceability across tools. This created significant strain on capacity and extended delivery timelines. While converted data was a critical component, the broader challenge was validating the overall solution under tight program timelines and governance expectations.
The question was clear: Could AI materially accelerate end‑to‑end solution validation while maintaining quality, control, and expert oversight?
Embedding AI into testing and validation execution
KPMG designed and delivered an AI automation testing pilot to determine whether AI‑assisted scenario generation and automated UI testing could reduce manual effort and accelerate testing cycles.
Using a secure, enterprise‑hosted large language model and structured prompt engineering, the solution ingested process maps and supporting documentation, parsed flows into structured natural‑language test scenarios, converted approved scenarios into executable automated UI test scripts, and triggered automated UI test execution with traceable outputs including screenshots, logs, and HTML reports.
Results were integrated with the organization’s existing testing and defect‑management tools to support end‑to‑end traceability.
AI did not replace subject matter experts. It accelerated the drafting and structuring of outputs. A human‑in‑the‑loop model was embedded at every stage, with business and QA SMEs reviewing scenarios, validating scripts, confirming selectors and controls, and approving execution results to ensure accuracy, completeness, and alignment with enterprise governance and risk controls.
The pilot evaluated AI‑generated test scenarios, AI‑generated Playwright scripts, and automated test execution within the target core platform environment to quantify effort reduction while preserving auditability and quality controls.
Accelerated, governed solution validation at scale
The pilot demonstrated measurable efficiency gains across end‑to‑end solution validation, including legacy data transformation and overall platform testing.
Testing scenario drafting time was reduced by approximately 50 per cent, while achieving approximately 85 per cent coverage compared to user‑written baselines. Playwright scripting effort was reduced by approximately 75 per cent, with strong alignment to human‑written logic and structure. Automated execution time was reduced from roughly 40 minutes manually to approximately one minute, with automated capture of execution evidence to support audit and governance requirements.
The result was not full automation. It was structured acceleration—AI‑enabled drafting, scripting, and execution combined with formal human review, governance, and traceability controls.
By embedding AI into both upstream conversion activities and downstream solution validation, the organization established a closed control loop where data‑transformation quality is continuously validated through system behaviour, not documentation alone.
How we make the difference
Successful transformation requires more than implementing new technology. It requires rethinking how change is delivered.
KPMG worked alongside this organization to identify high‑impact AI use cases within transformation execution itself, piloting practical applications that balanced speed, quality, and governance.
By embedding AI into testing and validation activities, KPMG helped demonstrate how transformation programs can move faster with confidence, preserving strong oversight while materially reducing delivery effort and timelines.
If you are exploring how AI can be embedded within transformation execution, not just within the target solution, KPMG can help.
KPMG. Make the Difference.
Meet the team
Michael Klubal
National Industry Leader, Infrastructure, Government and Healthcare
Toronto
KPMG Canada
