Leadership
Enterprise AI portfolio
In a large company the model is the small part. Around it sit the data it is allowed to see, the systems it has to integrate with, the evidence that it is good enough to trust, and whoever keeps it working a year from now.
My responsibility
I lead the Data & AI practice at NTT DATA with a team of 15, accountable for an AI solution end to end: choosing the problem worth solving, agreeing what counts as working, designing how it integrates and what it may see, funding it, and running it once it is live.
Scope
- Which problems are worth solving, and what counts as working before anything is built
- Data access, quality and the legal basis for using it
- Architecture, tooling and integration with the systems the process already runs on
- Acceptance criteria and evaluation agreed with the business before go-live
- Cost, ROI and total cost of ownership, including what inference costs at volume
- Life after go-live: monitoring, drift, retraining, support and rollback
- Schedule, dependencies and delivery risk across the programme
- The commercial model for AI services: offerings, pricing and deal structure