Engineering that spans cloud to AI.
A senior, hands-on consultancy covering the full path from cloud foundations to production AI — delivered as code, documented, and handed over cleanly.
Cloud & Solution Architecture
The architecture decisions you make early shape everything that follows. I design cloud foundations that are secure, cost-aware and built to scale — and I document the trade-offs so your team understands the "why", not just the "what".
- Well-Architected reviews and target-state architecture across AWS, Azure, GCP and OCI
- Landing zones, multi-account/subscription structure, networking and connectivity
- Identity, security baselines, guardrails and policy-as-code
- Cost modelling, FinOps practices and right-sizing
- Migration and modernisation strategy (lift-shift through to re-architecture)
Platform Engineering & Infrastructure as Code
Manual infrastructure does not scale and does not sleep. I build the automated, self-service platforms that let developers move fast without cutting corners — with Terraform at the core.
- Reusable Terraform modules, remote state and workspace strategy
- Internal developer platforms and golden paths
- Environment provisioning, drift detection and policy enforcement
- Secrets management with Vault / native secret stores
- Observability baked in: metrics, logs, traces and sensible alerting
Kubernetes, Containers & Delivery
Kubernetes is powerful and unforgiving. I deliver clusters and pipelines that are production-ready from the start — with the guardrails and automation that keep them that way.
- Production cluster design on EKS, AKS, GKE or OKE
- GitOps delivery with ArgoCD / Flux and Helm
- Autoscaling, resource governance and multi-tenancy
- CI/CD on GitHub Actions or Bitbucket Pipelines — build, test, scan, deploy
- Progressive delivery, rollbacks and release safety
AI, GenAI & MLOps Enablement
AI features fail in production for the same reasons software does — no automation, no observability, no evaluation. I apply cloud-engineering discipline to AI so your models and LLM features are dependable, measurable and cost-controlled.
- RAG assistants, LLM integrations and agentic workflows, built for production
- Evaluation harnesses, guardrails, prompt and model version control
- MLOps pipelines: training, deployment, monitoring and retraining
- GPU-aware infrastructure, vector databases and inference scaling
- Token-cost budgeting, caching, tracing and drift monitoring
Flexible, transparent ways to work together.
DeployCraft Ltd contracts on a B2B basis. Most engagements are assessed as Outside IR35, and I'm happy to work through a fair, role-based assessment for any contract.
Project delivery
A defined outcome — a platform, migration or AI capability — scoped, priced and delivered to an agreed definition of done.
Fractional / retained
Ongoing architecture and platform capacity, a set number of days per month, for teams that need senior input without a full-time hire.
Advisory & review
Architecture reviews, cost or security assessments, and second opinions — short, focused engagements with clear recommendations.
Not sure which fits?
Describe the problem and I'll suggest the most cost-effective way to tackle it.
Start a conversation →