How do engagements start?

Always with a 30-minute discovery call. We talk through your situation, share initial thoughts, and if it's a fit we scope the right engagement. Every call is a paid, focused session. You always get a clear quote first.

Do you work with my cloud or stack?

Almost certainly. We work across AWS, Azure and GCP, Kubernetes, Terraform and OpenTofu, plus the full CI/CD and observability ecosystem. If it runs in production, someone on the team has probably operated it.

How does payment work?

Fixed-scope work like the Power Hour and Audit is paid upfront via secure checkout. Retainers are invoiced monthly. You always get a clear quote before anything starts.

Can you work with my existing team?

That's the goal. We embed with your engineers, pair, document, and hand things over so your team grows stronger. We aim to make ourselves unnecessary, not indispensable.

Are you available for short or urgent work?

Yes. The Power Hour is perfect for a focused problem or a fast second opinion. For incidents or urgent reviews, mention it when you book and we'll prioritize.

Will any of this cause downtime?

No. Committed Use Discounts are a billing change with zero infrastructure impact. Rightsizing and Spot moves are staged with your team, tested, and rolled out with fallbacks. The goal is a smaller bill nobody in engineering even feels.

Do Committed Use Discounts lock me in?

Less than people fear when they are bought well. We ladder CUDs to your real usage and lean on sustained use discounts for the variable layer, so you capture deep discounts without betting on a workload that might change next quarter.

How are CUDs different from AWS or Azure commitments?

GCP gives you spend based and resource based CUDs, plus automatic sustained use discounts that reward steady running with no commitment at all. Used together they can beat a naive Reserved Instance approach, and we tune the mix to your usage.

What does it cost to work with you?

The cost review is a fixed, paid engagement, and it usually pays for itself many times over. Implementation and ongoing optimization are scoped after the review, once we both know the size of the prize. You always get a clear quote first.

Which Google Cloud services do you cover?

Compute Engine and GKE, Cloud Run, Cloud SQL and Spanner, BigQuery, Cloud Storage, Cloud CDN and the rest of the usual suspects. If it shows up on your bill, it is in scope.

GCP Cost & Rate Optimization

Cut your Google Cloud bill by 20 to 40 percent. Keep every ounce of reliability.

Most teams overpay for Google Cloud without realizing it. We find the waste, put Committed Use Discounts to work, move the right workloads to Spot, and set guardrails so the bill stops climbing. You keep the savings.

Book a cost review →See what I optimize

No lock-in. No autopilot black box. Just a senior engineer who has done this many times.

cost-review / summarysavings found

Annual run rate today

$910K

After optimization

$573K

Committed Use Discounts$186K

Rightsizing and custom machines$95K

Idle cleanup and storage$56K

20 to 40%

typical bill reduction

3 wks

to first savings

downtime required

15+ yrs

running cloud in production

Where the money goes

Three reasons your Google Cloud bill is bigger than it should be

Commitment anxiety

Teams stay on on-demand pricing because Committed Use Discounts feel like a trap. So they leave discounts worth up to 57 percent on the table to avoid lock-in that good planning would never cause.

Wrong machine shapes

Predefined machine types rarely match the workload, so you pay for vCPU or memory you never touch. Custom machine types and newer families can cut that waste, but nobody has time to rework it.

Oversized and idle

Instances picked months ago run larger than they need. Old disks, forgotten projects and dev environments run all weekend, and cold data sits in standard storage.

What I optimize

Every lever that moves the bill

Rate and usage, worked together. Cutting the price you pay per hour matters little if half those hours are waste, so I do both.

Committed Use Discounts

Spend based and resource based CUDs laddered to your real usage, stacked on top of automatic sustained use discounts, so you get deep discounts without a risky long term guess.

Rightsizing and custom machines

Match Compute Engine to what the workload actually uses with custom machine types and newer families, so you stop paying for vCPU and memory that sit idle.

Spot VMs where it fits

Batch jobs, CI runners and stateless services moved onto Spot VMs for up to 91 percent off, with safe fallbacks so the workloads that cannot take an interruption stay put.

Idle, storage and egress

Shut down what nobody uses, tier Cloud Storage to nearline and coldline, clean up orphaned disks and snapshots, and fix noisy cross region and internet egress.

GKE and autoscaling

GKE tuned with the cluster autoscaler, node auto-provisioning and sensible requests and limits, so nodes scale to the workload instead of sitting half empty.

Visibility and guardrails

Cost allocation by team and service with billing exports and BigQuery, budgets and anomaly alerts, so the savings stick and the next surprise bill gets caught early.

AcrossCompute EngineGKECloud SQLCloud RunBigQueryCloud StorageSpannerCloud CDN

How it works

From first look to lasting savings

Cost review

We read your bill and usage and show you exactly where the money is going and how much is recoverable.

Savings plan

A prioritized plan with the quick wins first, the dollar impact of each, and the risk spelled out plainly.

Put it in place

We do the work with your team in the loop. Commitments bought, resources tuned, guardrails set, nothing broken.

Keep it lean

Dashboards and alerts handed over so your team owns it, plus a monthly check in if you want ongoing eyes on it.

A recent engagement

34 percent off by fixing commitments and machine shapes

A data heavy startup ran Compute Engine and GKE on-demand with predefined machines far larger than the work required. We layered Committed Use Discounts, moved to custom machine types and tuned GKE, all with zero downtime.

We had no idea how much we were wasting on machine types that never fit. The savings landed almost immediately.

34%

lower monthly bill

$286K

saved in year one

90%

of compute on a discount

3 wks

to first savings

Questions

What people ask before we start

Most teams land between 20 and 40 percent of their Google Cloud bill. The exact number depends on how much you have optimized already. The cost review gives you a real figure for your project before you commit to anything bigger.

Also optimizingAWS cost optimization →Azure cost optimization →

See what your Google Cloud bill could look like.

Book a paid cost review. We read your account and show you the real number you could be saving, with a plan to get there. No autopilot, no lock-in, just a straight answer.

Book a cost review →Email Dieudonne