01
The situation
The company asked James to establish an enterprise data and analytics function from a standing start, for a multi-product business serving customers in more than 25 industries.
Case study · A global software company
From a standing start to production AI, in 50 governed stages
An enterprise data function built from nothing: strategy, warehouse, automated reporting and an AI finance assistant.
50 stageseach with a go or no-go decision
- Organisation
- A New Zealand-headquartered software company with 300+ staff, working in 70 countries
- Role
- Data and Analytics Manager, then Head of Data and Analytics
- Years
- 2023 to 2026
- Worked with
- The CEO, finance, commercial and revenue leaders, and an eight-director steering committee
02
What was built
He wrote and delivered the enterprise data strategy: 500 points across 50 stages, with governance, compliance and risk built in. The CEO adopted it, and an eight-director steering committee made a go or no-go decision at every stage on evidence of value delivered. Version 2.0 added agentic AI, semantic model governance and AI-specific controls.
Delivery included an enterprise Snowflake warehouse joining the finance, CRM, time-tracking, project and HR systems; a month-end automated performance pack with every KPI and OKR on one executive page and 622 self-auditing pages behind it; machine-learning revenue analytics; and a production AI application that answers finance questions from governed data in under 30 seconds.
03
The result
Leadership saw financial, operational and commercial performance together each month, from governed data instead of separate spreadsheets. The AI application was handed to a non-technical owner, with documentation written for that reader. A security project in the programme was the first at the company to score 100% on an external security review and penetration test.
James presented the data journey at the Snowflake Christchurch User Group in August 2024, and the integration work became a published customer story for Omnata, a Snowflake integration partner.
His support during the Snowflake programme, including a pivotal change in our Business Central ingestion approach, was instrumental to its success.
During a security project, his strategic insights and data analysis skills gave us exactly the information we needed, which helped us achieve the goal of the project without overspending and finishing it well within the deadline.
From recommendations on James’s LinkedIn profile.
This work was done by James Graham as an employee or contractor before Crucible was formed. It is described from public information, and it is not presented as Crucible client work.
Contact
Want this kind of result?
Tell us the decision you need to get right, and where the numbers are getting in the way.