Home / Insights
Insights
Architecture notes from YTJK Systems, and key public references for cloud and AI system design.
AI system designReference architecture: retrieval-augmented generation (RAG)
Ingestion and query paths for grounding language models in organisational content, with access control, evaluation and cost in view.
Read the note → Platform architectureDesigning a multi-tenant Kubernetes platformIsolation tiers, tenant onboarding as code and shared platform services for a Kubernetes platform used by many teams.
Read the note → AI platform architectureServing LLMs in production: latency, cost and reliabilityA model gateway, routing, caching and capacity choices that balance latency, cost and reliability for LLM workloads.
Read the note → AI platform architectureAI gateway: one control point for models, agents and toolsIdentity, policy, guardrails, routing, cost and audit for every model call and agent tool call, handled once in one layer.
Read the note → ReferencesReference libraryKey public frameworks and guidance for cloud architecture, Kubernetes platforms and AI system design.
Open the library →