Cloud & Solution Architecture
Well-Architected designs across AWS, Azure, GCP and OCI — landing zones, networking, security and cost models that stand up to scrutiny and scale.
UK Cloud & AI Engineering Consultancy
Cloud platforms that just work — and AI that ships to production.
DeployCraft is the limited company of a senior cloud & solution architect. I help teams design, automate and operate resilient infrastructure across AWS, Azure, GCP and OCI — and turn AI ideas into services you can actually run.
Delivering on
AWSAzureGoogle CloudOracle Cloud
17 years enterprise infrastructure through to cloud & AI
4 clouds AWS · Azure · GCP · OCI delivery experience
Outside IR35 engagements welcome, UK & remote
Cloud → AI same rigour applied to MLOps & GenAI
What I do
Two disciplines, one standard of engineering.
A proven cloud and platform practice as the foundation — with AI and MLOps as a fast-growing specialism built on the same principles: automated, observable and reproducible.
Platform Engineering & IaC
Repeatable infrastructure with Terraform, reusable modules and golden paths so teams provision safely and ship without waiting on tickets.
Kubernetes & Containers
Production-grade clusters, GitOps delivery, autoscaling and observability — plus the guardrails to run them without 3am surprises.
CI/CD & Delivery Automation
Pipelines on GitHub Actions or Bitbucket that build, test, scan and deploy with confidence — trunk-based, automated and audit-friendly.
AI & GenAI Enablement
From RAG assistants to LLM integrations — production patterns, evaluation and cost controls that turn AI prototypes into dependable services.
MLOps & AI Platforms
The plumbing behind reliable AI: model pipelines, vector stores, GPU-aware infrastructure, and monitoring so models stay accurate and observable.
The AI service line
AI is only useful when it runs reliably.
Everyone can build a demo. Fewer can put a large-language-model feature into production with evaluation, guardrails, cost controls and monitoring. DeployCraft brings cloud-engineering discipline to AI — so your RAG assistant, agent or model pipeline behaves the same on day 100 as day 1.
- Retrieval-augmented generation & LLM integration patterns
- Evaluation harnesses, guardrails and prompt/version control
- GPU-aware infrastructure, vector stores and inference scaling
- Token-cost budgeting, caching and observability
How I work
Senior, hands-on, and easy to work with.
A straightforward engagement model designed to reduce risk and leave your team stronger than it found me.
01
Discover
Understand the goal, constraints and current state — architecture review, risk map and a clear definition of done.
02
Design
A pragmatic target architecture and delivery plan, sized to your team and budget, with trade-offs made explicit.
03
Build
Infrastructure as code, pipelines and platform components delivered in small, reviewable increments.
04
Enable
Documentation, runbooks and knowledge transfer so your team owns it confidently after I roll off.
Tooling
A modern, cloud-native stack
AWSAzureGCPOCITerraformKubernetesDockerHelmGitHub ActionsBitbucketArgoCDPythonPrometheusGrafanaVaultLangChain
Have a platform to build or an AI idea to ship?
Tell me what you're working on. I'll come back with an honest view on approach, effort and where I can add the most value.