STADLER reshapes knowledge work at a 230-year-old company
Key Points
- 125+ custom GPTs deployed
- 30–40% time saved on routine tasks
- >85% daily active usage
Summary
STADLER embedded ChatGPT across ~650 employees to convert hours of knowledge work into minutes, creating a company-wide AI productivity layer. Results include 125+ custom GPTs, 30–40% time savings on common knowledge tasks, 2.5x faster time-to-first-draft on average, and >85% daily active usage. Adoption combined bottom-up experimentation with top-down access, training, and guardrails, and spans engineering, project management, marketing, translations, and general drafting and summarization.
Key Points
- Deployment model: enable experimentation while providing company-wide access, training, and governance.
- Scope: integrated across functions—engineering (analysis, code support), project teams (process templates, docs), marketing (translation, content), and company-wide drafting/summarization.
- Measured impact: 30–40% time savings on routine knowledge tasks; 2.5x faster first drafts (up to 6x in high-volume cases); >85% daily active usage.
- Deliverables: 125+ custom GPTs focused on translation and email workflows, plus many team-specific assistants.
- Strategic direction: move from assistant to execution layer—agents that gather info, generate and validate outputs, and route work for approvals.
Implementation notes (practical for engineers)
- Start by identifying high-frequency, repeatable tasks (translation, email, first drafts, summaries) and build focused custom GPTs.
- Pair bottom-up pilots with central enablement: provide templates, training, and clear guardrails for safety and compliance.
- Measure impact with metrics like time-to-first-draft, percent time saved per task, and daily active usage.
- Plan next-phase integrations: build agentized workflows that connect data sources, run validations, and automate routing/approvals.
- Iterate on prompts, tooling, and observability to keep outputs consistent and auditable.