Been running the numbers on our AI/ML stack for the last year. The infrastructure savings from using Spot and Graviton are great... but then I saw the line item for "Prompt Engineering Training Bootcamp" 😳
It feels like we're just swapping one cost center for another. To get real ROI, you need:
* Specialized, expensive courses for teams
* Ongoing "prompt tuning" hours that are hard to track
* The churn cost when a good prompt engineer leaves
Our rough breakdown for a team of 5:
```python
# Annual Costs (est.)
infra_savings_vs_on_prem = 120000
training_budget = 25000 # courses, workshops, certs
productivity_hours_lost = 15000 # learning curve
total_tco_impact = infra_savings - training_budget - productivity_hours_lost
# That's a $80k net... not the $120k we projected.
```
Anyone else seeing this? How are you baking these human costs into your FinOps models for GenAI?
#savings
I totally get this. We've started treating prompt engineering more like data quality work, building reusable templates and a small internal library. It cuts down on the "ongoing tuning hours" you mentioned because teams aren't starting from scratch each time.
Have you considered rolling that training in-house? We found a few senior folks who learned the ropes and now do short, focused sessions. It's way cheaper than bootcamps, though you're right about the churn risk if they leave.