Topic
DevOps
Technical insights from FlareTech about DevOps.
- Method12 min read
Feature flags for a small team — without the platform theatre
A feature flag is a runtime decision. Not a dashboard subscription, and not an env var you forget to turn off.
→Feature flags for a small team — without the platform theatre - Method12 min read
How to run an honest discovery week
A discovery week is five days that replace the requirements spec nobody can estimate from — with decisions you can actually build.
→How to run an honest discovery week - Cloud10 min read
EU hosting: what it actually means for compliance
An EU region is not the same as GDPR compliance. What counts is where the data lives, who can read it, and which subprocessors you have actually accepted.
→EU hosting: what it actually means for compliance - Cloud8 min read
Why your cloud bill keeps rising — and what to do
Cloud bills rarely rise only because you grew. They rise because egress, idle resources and unclear ownership are allowed to accumulate — until someone finally opens the invoice.
→Why your cloud bill keeps rising — and what to do - Cloud5 min read
When is AWS the right call — and when is it too much?
AWS can carry almost anything — but "almost anything" is not the same as "what you need right now". Here is how we choose between AWS, Vercel, and Cloudflare without cloud religion.
→When is AWS the right call — and when is it too much? - Method7 min read
Infrastructure as Code: when it is actually worth setting up
Terraform is not free — it has its own learning curve and its own maintenance. Here is the honest threshold for when infrastructure as code pays off, and when clicking around in a cloud console is still the right call.
→Infrastructure as Code: when it is actually worth setting up - Method7 min read
Observability without theatre: what actually gets looked at when something breaks
A dashboard with forty panels that nobody opens is not observability — it is decoration. Here is how we build logging, metrics, and alerting that someone actually looks at during an incident, instead of afterwards.
→Observability without theatre: what actually gets looked at when something breaks - Method6 min read
CI/CD with preview deploys — why it pays off
A pull request without a clickable preview is a review in the dark. Preview deploys make code review visual, catch bugs before they hit production, and give stakeholders something concrete to react to — without running the project locally.
→CI/CD with preview deploys — why it pays off - Cloud4 min read
Cloudflare for production apps — what it gives you, and what it doesn't
Cloudflare is more than a CDN. For many Next.js and SaaS products it's edge, DNS, WAF, storage, and compute in one vendor — but you need to know which pieces you actually need.
→Cloudflare for production apps — what it gives you, and what it doesn't