Solution Design & Architecture

Translate business requirements into Claude solution architectures and choose the right pattern.

1. 1.1From Requirements to ArchitectureHow to turn an ambiguous business request into a defensible Claude solution architecture, using the questions, artefacts and decision gates a professional architect is expected to apply.9 min
2. 1.2Workflows vs Agents: Choosing the Right Level of AutonomyA decision framework for choosing between code-orchestrated workflows and model-directed agents, with the production consequences of each choice.9 min
3. 1.3Reference Architectures for Claude SolutionsThe recurring production shapes for Claude systems, the platform features each one leans on, and the hosting and runtime choices that sit underneath them.10 min
4. 1.4Designing Multi-Agent Systems at ScaleWhen multiple Claude agents genuinely beat one, how to structure orchestration and communication, and how to keep cost, correctness and observability under control.10 min
5. 1.5Data, Memory and State DesignWhere state lives in a Claude solution, how to manage context over long horizons with compaction, context editing and memory, and how to design data flows that respect classification and retention.10 min
6. 1.6Cost, Latency and Quality Trade-offsThe levers an architect has over cost, latency and quality in a Claude system, the order in which to pull them, and how to reason about cost per completed task rather than cost per request.10 min