Getting AI to deliver in the life sciences sector requires more than good models. One pharma company shares how it's building the data infrastructure that makes AI usable in practice, including lessons learned from bridging the scientific and engineering sides of the work.