V1.0Marco 1 — Core Platform Foundations
ACCEPTED2024-03-10
ADR-001: GitOps over imperative pipelines
Use declarative reconciliation from Git instead of imperative deployment pipelines to drive cluster state.
100% Concluído (4 fases)
ADRs & Roadmap
Decisões de Arquitetura & Roadmap.
Contexto, trade-offs, decisões e marcos de evolução implementados e planejados para a plataforma.
Launch stateThese entries are placeholders while the real case studies and experiment write-ups are prepared for publication.
Decisões Registradas
6
ADR-001 até ADR-006Aceitas & Implementadas
5
Em produção no siteEm Progresso / Propostas
1
ADR-006 (Busca Híbrida)Conclusão dos Marcos
94%
Evolução do Sistema
Trajetória de decisões de infraestrutura, IA nativa, visualização 3D e retrieval semântico.
V1.3Marco 2 — Grounded Operational Knowledge
ACCEPTED2025-07-10
ADR-002: Retrieval over fine-tuning for runbook answers
Answer operational questions by retrieving from runbooks rather than fine-tuning a model on internal incidents.
100% Concluído (3 fases)
V2.0Marco 3 — Native AI Engine Migration
ACCEPTED2026-09-06
ADR-003: Google Gemini API native integration for Ask Alan RAG
Adopt Google Gemini 2.5 Flash via the official @google/genai SDK for Ask Alan RAG generation, replacing proprietary proxies.
100% Concluído (4 fases)
V3.0Marco 4 — Living Brain 3D Spatial Knowledge
ACCEPTED2026-09-06
ADR-004: Deterministic 3D force-directed knowledge graph
Use Three.js and deterministic pseudo-random force layout for the 3D living brain instead of dynamic unseeded physics.
100% Concluído (4 fases)
V3.1Marco 5 — Concentric Spatial Hierarchy
ACCEPTED2026-09-06
ADR-005: Semantic layer depth partitioning for complex graphs
Partition the 3D knowledge brain into 3 progressive semantic layers: Core, Practice, and Foundations.
100% Concluído (4 fases)
V2.5Marco 6 — Semantic Brain & Hybrid Retrieval
PROPOSED2026-09-06
ADR-006: Vector embeddings and hybrid search for knowledge retrieval
Adopt hybrid retrieval combining lexical BM25 matching with vector embeddings (text-embedding-004) for Ask Alan.
65% Concluído (4 fases)
- Standardization on text-embedding-004Concluído
Selected Google text-embedding-004 (768 dimensions) for superior semantic understanding across cloud native and DevSecOps domains.
- Reciprocal Rank Fusion (RRF)Em desenvolvimento
Implemented fusion ranking combining lexical keyword BM25 candidates with semantic cosine similarity scores before LLM generation.
- PostgreSQL pgvector Schema MigrationPlanejado
Provisioning pgvector table with automated re-indexing triggers executing whenever content is committed via /write.
- Semantic Recall & Grounding EvaluationPlanejado
Evaluation suite testing conceptual paraphrase queries (e.g. 'zero trust ingress' vs 'Istio mutual TLS') to ensure 100% citation grounding.