Infrastructure that stays up
PathOr designs, builds and operates the infrastructure applications run on. Cloud architecture, deployment pipelines, scaling and the monitoring that tells you something is wrong before your customers do.
Cloud architecture
AWS and Azure environments designed around your actual workload.
Containers
Docker and Kubernetes, sized to the problem rather than the trend.
CI/CD
Deployment that is boring, repeatable and reversible.
Auto-scaling
Capacity that follows demand instead of guessing at it.
What we do
Cloud architecture on AWS and Azure — designing the environment an application lives in, including networking, storage, security boundaries and cost shape. The last of those gets skipped more often than it should; infrastructure that works but costs three times what it needs to is still a problem you have to solve later.
Containerisation and orchestration with Docker and Kubernetes, where the workload justifies it. Continuous integration and deployment pipelines so releases stop being events. Auto-scaling, performance optimisation and the observability to know whether any of it is working.
Deployment should be boring
The goal of a good pipeline is that shipping stops being interesting. A deploy should be repeatable, reversible and frequent enough that no single one carries much risk. Where releases are rare and manual, every one is an event, and events are where outages come from.
That usually means fixing the pipeline before anything else. It is unglamorous work and it pays back immediately.
Running what we build
PathOr operates its own products on this infrastructure — AvA and PathOr Meet both run on it, with real users and real uptime expectations. Meet in particular is a live video service, which is an unforgiving thing to keep running.
Building infrastructure you then have to answer for is different from building it for handover. It makes you conservative in the right places.
Common questions
Which cloud providers does PathOr work with?
+
Primarily AWS and Azure. Which one suits a given project depends on the workload, existing commitments and cost profile rather than a fixed preference.
Does PathOr offer ongoing infrastructure support?
+
Yes. Alongside initial architecture and migration work, PathOr takes on ongoing operation, monitoring and optimisation of the infrastructure it builds.
Do we need Kubernetes?
+
Often not. Kubernetes solves real problems at a certain scale and adds real operational cost below it. Part of the scoping work is being honest about which side of that line a workload sits on.
Infrastructure giving you trouble?
Whether it is cost, uptime or deploys that everyone dreads, tell us what is happening and we will tell you what we would do about it.
Start a conversation