Build value.
Keep costs in control.
Practical insights connecting technology choices with financial decisions. Cloud, AI and governance: one brick at a time.
Explore the insights ↓
Beyond the LinkedIn post.
Context, tradeoffs and practical next steps, in French and English.
Tokens are billed, value is measured elsewhere: structuring generative AI cost tracking
Dashboards that only display millions of tokens cannot explain spending or support trade-offs. This article details a three-level approach — visibility, attribution, decision — before any optimization, and offers a practical checklist inspired by Jérôme Raguillet's original post.
EBS snapshots: when archiving increases the bill instead of reducing it
An AWS case with 94,000 snapshots and 1.9 PB of referenced data shows that the decision criterion is not a snapshot's age, but its role in the restore chain and its full cost after archiving. An analysis of the mechanics, the triage reasoning, and the safeguards to put in place.
Tokenomicon and FinOps X Amsterdam: what exploding AI costs change for FinOps
In his post about the Amsterdam events, Jérôme Raguillet offers a reading of the emergence of Tokenomics: dramatic token consumption growth, hidden costs missed by most dashboards, and early tangible standards. A framing to help FinOps teams decide where to focus their efforts in 2026.
FinOps and AI: simulating token costs before the bill — what Infracost AI announces
After years in which engineers only discovered cloud costs once in production, Infracost has announced Infracost AI, currently in early access, to bring together AI consumption and simulate the impact of model and routing choices. This article examines the promise, the points to verify, and the questions to ask before drawing conclusions.
S3 Intelligent-Tiering: escaping the manual tiering trap
Manually classifying data based on an access assumption exposes you to costly discoveries, such as a quarterly report that re-reads data deemed cold. This article, based on a LinkedIn post by Jérôme Raguillet, details how S3 Intelligent-Tiering's tiers work, the decision criteria, the migration approach and the watch points, with a costed example drawn from the original post.
R9g and R9gd on AWS: what the new Graviton5 generation really changes for your FinOps strategy
AWS has announced R9g and R9gd instances, built on Graviton5, with up to 25% more compute performance than R8g/R8gd. Beyond the technical announcement, the real issue for a FinOps team is the impact on existing commitments: Compute Savings Plans, EC2 Instance Savings Plans and Reserved Instances do not react the same way to a generation change. This article walks through the reasoning, the decision criteria and a phased migration approach.
Cost is a starting point.
Value is the destination.
I help connect Cloud, AI and IT financial governance decisions. My approach: make costs understandable, clarify ownership and support informed tradeoffs.
Let’s talk on LinkedIn ↗- 01Understand usage
- 02Inform decisions
- 03Manage value