How We Build Platforms
Our HPC & Infrastructure Process
A documentation-first approach to building and automating HPC and Linux platforms, from requirements through to production handover.
This is a general outline. The exact phases and deliverables for your project are agreed during discovery and set out in your Scope of Work.
HPC & Infrastructure Journey at a Glance
01
Requirements
To start
02
Architecture
Before build
03
Automation
Automation first
04
Build
Main build
05
Testing
Before handover
06
Handover
Project close
07
Ongoing
After handover
Full Transparency
What You Can See at Every Stage
No black boxes. You can see your assets, designs, and project progress from day one.
Your Project Workspace
Requirements, architecture, runbooks, and test results in one place.
Architecture Documents
Full diagrams shared before we build. Nothing is built until you approve the design.
Monitoring Dashboard
Once live, you can see uptime, utilisation, and alerts yourself.
Automation Repository
Every change is code in version control — you can see exactly what was configured.
Progress Updates
Written updates through the build: what's configured, what's next, what access we need.
You Own Everything
Code, configuration, credentials, and documentation belong to you. Full handover at close.
Phase by Phase
Every Step, Explained
What happens at each stage, and what we need from you.
Phase 01 · Requirements
To start
Understanding Your Workloads
We start with technical discovery: the jobs you run, how they scale, your storage and access patterns, and any compliance constraints.
What you do
- Tell us what workloads you need to run
- Share benchmarks or job profiles, if you have them
- Flag any compliance requirements
What we deliver
- A written requirements specification
- Workload and capacity analysis
- An initial platform recommendation
Phase 02 · Architecture
Before build
Designing the Platform
We design the cluster: scheduler and queue policy, storage layout, the access model, and how it recovers from failure.
What you do
- Review the design documents
- Confirm which servers or cloud accounts we build on
- Sign off before we start building
What we deliver
- Logical architecture and network diagrams
- Scheduler, queue, and storage design
- Resilience and disaster recovery design
Phase 03 · Automation
Automation first
Provisioning as Code
Everything is provisioned with Ansible rather than by hand, so the build is repeatable, reviewable, and can be rebuilt from scratch later.
What you do
- Give us access to the target machines or cloud accounts
- Confirm naming, network, and access conventions
- Review the automation repository
What we deliver
- Ansible roles and playbooks covering the whole build
- Version-controlled configuration that you keep
- A build you can run again from scratch
Phase 04 · Build
Main build
Cluster & Storage Build
We provision the operating system, scheduler, shared storage, and user access — hardening as we go and recording every change.
What you do
- Review each part of the build as it lands
- Test job submission and file access
- Share credentials for systems we integrate with
What we deliver
- Linux provisioning and security hardening
- Scheduler and queue configuration
- Shared storage, user access, and job submission
Phase 05 · Testing
Before handover
Validation & Benchmarking
Before handover we benchmark performance, test failover, and check the hardening — and share the results.
What you do
- Review the test plan
- Provide representative workloads
- Confirm the acceptance criteria are met
What we deliver
- Performance benchmarks against agreed criteria
- Failover and recovery test results
- A security hardening report
Phase 06 · Handover
Project close
Documentation & Training
A system your team can't operate is a liability. We document everything and train your team before we close the project.
What you do
- Nominate people for training
- Attend the operations walkthrough
- Review and sign off the documentation
What we deliver
- An operations runbook
- Disaster recovery and monitoring documentation
- A training session, on site or remote
Phase 07 · Ongoing
After handover
Monitoring & Support
Platforms drift. We offer monitoring and support so problems are caught before they become outages.
What you do
- Define escalation contacts
- Review the performance reports
- Tell us about planned capacity changes
What we deliver
- 24/7 monitoring and alerting
- Regular uptime and performance reporting
- OS patching and package updates
Our Standards
How We Operate
Design before build
Nothing is built until the architecture is signed off. Changes at design stage are free.
Automate, don't click
Every change is code in version control, so the platform can be rebuilt from scratch.
Validation before handover
The system doesn't pass to you until it meets the agreed acceptance criteria.
Training is not optional
We're not done until your team can run the system without us.
Ready to Start?
Ready to Build Your Platform?
Tell us what you need to run. We'll be honest about what it takes to build and operate it.
Book a Discovery CallExplore Our Other Services
We cover website development, app development, social media marketing, and enterprise HPC infrastructure.
View All Services