Azure FinOps Essentials

Hi there,

It has been a while since the last edition went out. The silence was not intentional, but it does deserve an explanation.

Over the past months, I made a significant career move. I transitioned from working as an Azure Cloud consultant at Xebia to stepping into the role of CTO at Freelance.nl. It is an exciting new chapter, with broader responsibilities and a different focus. We still use Azure. FinOps still matters. But I am no longer helping customers with their cloud cost challenges on a daily basis.

With that shift came a reality check. I simply do not have the time and energy to continue publishing this newsletter twice a month at the level it deserves. As much as I enjoyed writing about Azure FinOps, this will likely be the final edition of Azure FinOps Essentials in its current form.

Before closing this chapter, I want to share something I have been building. It combines the topics I care most about: cloud, FinOps, shift left, and GitHub. A platform designed to bring FinOps closer to developers.

And to make it interesting, the first 10 people who want to try it will receive a Pro account for free.

Let me show you what I mean.

Bringing FinOps where it belongs

If you’ve been reading this newsletter for a while, you know I’ve repeated one theme often.

Shift left.

The closer you bring an issue to the people who can act on it, the faster it gets resolved. That applies to performance, security, and just as much to cloud cost.

The problem with most FinOps tooling is not that it lacks insight. It is that the insight lives somewhere else. In dashboards. In reports. In finance meetings. Far from the engineers who can actually fix the issue.

That gap is what I wanted to close.

So I built LeftSize.

LeftSize runs from your own GitHub repository using GitHub Actions. You connect your Azure subscriptions or AWS accounts, and it scans them for cost optimizations, security risks, governance gaps, deprecated components, and best practice violations.

Instead of showing you a dashboard, it creates GitHub Issues.

Each finding becomes an actionable item inside your existing workflow. Engineers can assign it, discuss it, generate remediation scripts through ChatOps, snooze it, or fix it. No context switching. No separate portal.

It brings cost and cloud control directly into engineering.

This is not built for enterprises with a 10 million Azure annual spend and a dedicated FinOps team. There are strong platforms for that market.

LeftSize is for smaller and growing teams that want grip and control without implementing a heavy FinOps stack. It is easy to onboard, runs in your own pipeline, and turns cloud insights into actual work.

And to make it interesting, the first 10 people who want to try it will receive a Pro account for free.

If you believe cost awareness belongs closer to code, I would love for you to try it.

How it works in practice

LeftSize is intentionally simple to set up.

You install the GitHub App in your organization.

You add a workflow file to your repository.

That workflow runs as a GitHub Action and connects to your Azure subscriptions or AWS accounts using read-only credentials.

Nothing exotic. No long onboarding. No external credential storage. You stay in control because everything runs inside your own GitHub pipeline.

If you manage multiple subscriptions or accounts, you can use a matrix workflow. That allows you to scan several environments in parallel. Production and non-production can be separated and treated differently.

Once configured, the workflow runs on a schedule. Daily. Weekly. Or however often you prefer.

LeftSize scans your environments and looks for cost optimization opportunities, governance gaps, security risks, deprecated components, and configuration best practices. There are around 100+ rules it scans for; on AWS and Azure.

When it finds something, it does not send you to a dashboard.

It creates a GitHub Issue.

LeftSize understands environments. Production findings are treated differently from non-production ones. Findings are grouped to avoid noise. Snoozed items remain snoozed. Resolved issues stay resolved. The goal is not to spam you. The goal is to create manageable, actionable work.

From there, you are in control.

You triage the issue.
You look at the estimated impact.
You check the context.
You use ChatOps to ask for remediation scripts or deeper explanation.
You request an AI prompt to investigate further.

Or you decide to accept the risk and ignore it.

Labels help you prioritize based on impact and effort. High savings potential. Security critical. Governance improvement. Or low priority.

Everything happens inside the workflow your engineers already use.

No separate portal.

No new process to enforce.

Just actionable insights, delivered where work already happens.

Closing this chapter

For now, this is where I pause.

I will not continue sending regular Azure FinOps editions. At least not in this format. If I ever return to practicing FinOps on Azure in the same way as before, and have meaningful lessons to share, you will hear from me again. But I do not want to promise something I cannot consistently deliver.

What I do hope is that you stay curious about cloud cost. That you keep challenging defaults. That you keep bringing responsibility closer to engineering.

And I would genuinely appreciate it if you gave LeftSize a try.

If you are interested, send a short email to [email protected] or simply reply to this email.

The first 10 people who reach out will receive a Pro license, so you can explore the full platform without limitations.

Thank you for reading.
Thank you for the conversations.
Thank you for being part of this journey.

Michiel