Laura Keller
Product Manager · Product
LK
ROI score
47Watch
Cost per accepted unit is above the team median.
Recommendation
Compare this person's model mix to top performers on the team.
AI spend
€292
2% vs prior period
Yield rate
43%
Team average 61%
Cost per accepted unit
€13.64
Team average €8.83
Leverage vs last month
1.17
50 runs · 21 accepted · 4.3M tokens
Weekly AI spend
Human loops
Accepted46%
Rejected 54% · team 72% accepted
Guidance levelHigh
Runs without result7%
Waste€178
ROI score breakdown
Yield 30%
43
Cost efficiency 25%
28
Trend 20%
61
Model fit 10%
91
Capability use 15%
35
Each component is scored 0–100; 50 means on par with the team median or flat vs last month. Good ≥ 70, Watch 40–69, Bad < 40.
Model mix: spend share vs token share
Premium reasoning model32% spend · 11% tokens
Standard model51% spend · 41% tokens
Fast small model17% spend · 48% tokens
Top skills used
- Code assist
- Data analysis
- Meeting notes
Agents and MCP usage
- Calendar MCP40 runs
- Docs MCP194 runs