CloudBoostUP

CloudBoostUP is now listed in the Microsoft Marketplace partner directory. A small milestone for the team.

New blog: how to automate multicloud security change cycles without losing the human approval gate. https://lnkd.in/db6DgWq9

New case study on our site: how we migrated cloudboostup.com DNS to Azure DNS as Infrastructure as Code. https://lnkd.in/dksrgrUA

We just finished a Databricks data platform for a client. Unity Catalog, isolated data products, full governance stack. Then came a set of lightweight integration APIs. The obvious move was to run them on Databricks too. We ran the numbers first. 🔢 The cost difference was huge. 📉 We chose the cheaper option and deliberately stepped outside the central governance model. Here is why that was the right call, and what the trade-off actually looked like in practice. 🔧 🔗 [link in comments]

We scaled a data platform to tens of isolated products. Each business unit gets their own storage, compute, catalog, and identity. Fully isolated. Nothing shared. Not by writing tens of configurations. By writing one Terraform module and executing it through config.

We built a cloud platform from scratch for a Swiss enterprise. Not because we planned to. Because the existing setup wouldn’t let us move at engineering speed. So we did it our way - everything as code, every change auditable, zero portal clicks. That decision unlocked everything that came after.

We added a "From Our Work" section to cloudboostup.com. Not case studies with marketing polish. Detailed write-ups of real engagements — what we found, what we built, and what changed. Three are live now: → A cloud platform serving hundreds of teams at enterprise scale — self-service provisioning, 200+ governance policies, fully integrated with ServiceNow and Azure DevOps → A data platform advisory for a Swiss startup — no code deployed, just infrastructure analysis, architecture decisions, and an execution-ready backlog → Multi-environment lifecycle automation for a finance transformation — production replicas on demand, 3–12 hours fully automated

We wrapped a multi-year cloud infrastructure engagement at the end of 2025. Here’s what we delivered: → Full multi-environment lifecycle automation — from dev to production, fully reproducible → Every resource provisioned as code — no clickops, no undocumented changes → CI/CD pipelines with automated testing and drift detection → Policy-as-code guardrails — misconfigurations caught before they reach production → Handover complete — the client’s team owns and extends it independently A highlight along the way: the work contributed to a top-3 finish at the IDC Future Enterprise Awards 2023 — Best in Future of Digital Infrastructure, Czech Republic. Infrastructure done right doesn’t just run well. It earns recognition. Full write-up below. 👇

If your lead engineer left tomorrow, how long would it take to rebuild your Databricks workspace from scratch? Most teams we speak to pause at that question. The workspace was spun up by whoever had portal access at the time. Cluster configs exist in someone's head. Networking "just works" — nobody's entirely sure why. Storage grew one container at a time. That's not a Databricks problem. That's infrastructure treated as an afterthought. Here's what treating it as a first-class concern looks like: Workspace as code — Terraform provisions everything. You can tear it down and recreate it exactly. No clickops, no tribal knowledge. Private networking — VNet injection, private endpoints. Data doesn't touch the public internet. Non-negotiable in regulated industries. Unity Catalog from day one — not bolted on after six months of ACL spaghetti. One governance layer for data, notebooks, and AI assets. Cluster policies enforced by the platform — not by hoping engineers remember to terminate. Cost control isn't a policy document; it's a constraint. Infrastructure changes via Git — every workspace change is a PR. You get a history, a reviewer, and a rollback path. None of this is exotic. But almost every team that calls us skipped it — because it slowed down the PoC. We wrote up what production-ready Databricks infrastructure actually involves. ↓ Link in first comment

Most companies start their cloud journey by clicking around the portal. It works — until it doesn't. At some point you have 15 subscriptions, no naming standard, 3 people who "know how it works", and a security audit coming. That's when governance becomes urgent instead of important. Here's what a proper Cloud Landing Zone delivers: → Subscription structure designed for scale → Identity & access — least privilege, enforced by code → Policy guardrails — prevent misconfigurations before production → Everything in Git — auditable, reproducible, no clickops The shift isn't "add more rules." It's "make the right thing the easy thing." We wrote up what this looks like in practice 👇 (link in first comment)