Subsidiary of an electrification multinational (R&D area)

From public models on personal phones to an authorized internal environment

Challenge
  • R&D area with no AI strategy or owner.
  • Shadow AI: free LLMs from personal phones, with no criteria for writing prompts or evaluating answers.
  • Generational resistance; IT with no security policy.
Action
  • Maturity diagnosis and AI alignment workshop.
  • Prioritization of processes suited to automation.
  • Separate teams for those who design prompts and agents and those who evaluate them.
Result
  • Approved roadmap of 12 processes prioritized by ROI, benefit and effort.
  • Authorized internal LLM environment, built with IT, which replaced the use of public models.
  • AI integrated into the competency matrix; maturity from level 1 to 3.
Scope and status

Scope: one R&D area. Governance case, no financial figure. Client anonymized.

Other cases

See all cases →

Let's start with a viability assessment

Free 60-minute Discovery. We move forward if there is an executive with authority and budget to decide, and a designated internal owner.