AIDLC reliability dual-axis as plugins — Ontology Engineering (correctness) and Harness Engineering (safety) on AWS, with AIDLC Workflows as the process spine and AgenticOps closing the living-ontology Outer Loop.
AWS DevOps Agent for Claude Code: incident investigation, cost optimization, architecture review, topology mapping, and multi-AgentSpace routing.
Build and operate agentic AI workloads on EKS with vLLM, Inference Gateway, Langfuse, and Kagent. Covers GPU planning, model serving, routing, observability, and guardrails.
Agentic operations layer for AIDLC — self-improving loops, autonomous deploys, continuous evaluation, incident response, and cost governance.
AIDLC Phase 1 (Inception) + Phase 2 (Construction) opt-in extensions for awslabs/aidlc-workflows. Inception covers workspace detection, adaptive-depth requirements analysis, user stories, and workflow planning. Construction covers component design, code generation, TDD-for-agentic test strategy, risk discovery, and phase quality gates.
Legacy workload modernization to AWS — 6R strategy, workload assessment, to-be architecture, containerization, and cutover planning. Adapted from aws-samples/sample-ai-driven-modernization-with-kiro (MIT-0).