Where it started
A pharma device department had grown a collection of spreadsheets, scripts and macros over ten years: tolerance calculations, test data processing, requirement tracing. Engineers had built them to get the work done. Several were used in design verification. Nobody owned them, and some ran on one person’s laptop.
The strategic question
Which tools create enough value to be made professional, owned, validated where needed and strengthened with AI? Which should be merged, and which retired?
What we found
An inventory of 23 tools, each scored on hours saved per year, number of users, what happens if it is wrong, and the effort to make it professional. Six carried most of the value. Four were used in the design history file without any validation, which was the real risk. Nine duplicated each other or a system the company already paid for.