Markdown Editor
Mở trình soạn thảo

Bộ mẫu kiến trúc nền tảng AI doanh nghiệp

Thiết kế nền tảng AI doanh nghiệp: kiến trúc tham chiếu, tích hợp ứng dụng, LLM gateway, RAG, thiết kế đa agent, đánh giá và quản trị.

8 mẫu trong 4 bước, theo thứ tự sử dụng. Mở mẫu bất kỳ trong trình soạn thảo, điền nội dung rồi xuất ra PDF, DOCX hoặc HTML.

Bước 1

Chiến lược

  1. Enterprise AI Platform Reference Architecture Template

    Đầy đủ

    Capability map, LLM gateway, permissioned retrieval, agent runtime (MCP, A2A), evals, guardrails, telemetry, showback, 42001 and AI Act hooks

    MCP 2026-07-28; A2A 1.0.1; ISO/IEC 42001:2023; EU AI Act as amended by 2026/1744

  2. GenAI Enterprise Application Integration Design Template (ERP, CRM, ITSM)

    Đầy đủ

    Suite use cases, native vs custom scoring, vendor AI baseline, grounding and oversharing, integration, oversight, residency, evals, cost, AI Act

    Vendor documentation (SAP Joule, Salesforce Agentforce, Microsoft 365 Copilot, ServiceNow Otto); EU AI Act as amended by 2026/1744
Bước 2

Thiết kế

  1. LLM Gateway System Design Template

    Đầy đủ

    LLM gateway and router design: unified API, provider adapters, token quotas, routing and fallback, caching, PII redaction, guardrail hooks, metering

    OpenTelemetry GenAI semconv (Development); OWASP LLM Top 10 2026; RFC 9457

  2. RAG System Design Template (Retrieval-Augmented Generation)

    Đầy đủ

    Design a RAG system: ingestion, chunking, embeddings, hybrid search, citations, evaluation, ACLs, security and cost

    Common RAG practice; OWASP Top 10 for LLM Applications 2026

  3. Multi-Agent System Design Template (A2A, MCP, OWASP Agentic)

    Đầy đủ

    Multi-agent design: roster, topology, handoff contracts, shared memory, A2A and MCP links, budgets, cascading-failure controls, oversight, OWASP ASI mapping

    OWASP Top 10 for Agentic Applications 2026; A2A 1.0.1; MCP 2026-07-28
Bước 3

Đánh giá Well-Architected

  1. Generative AI Well-Architected Review Template

    Đầy đủ

    Workload context, responsible AI and agentic preamble, six-pillar GenAI questions mapped to AWS GenAI Lens, findings register, improvement plan, re-review

    AWS Well-Architected Generative AI Lens (Nov 2025); Azure WAF AI workloads; Google Cloud WAF AI and ML perspective

  2. Agentic AI Threat Model Template (OWASP ASI Top 10, MITRE ATLAS)

    Đầy đủ

    Agent trust boundaries, assets, autonomy, ASI01-ASI10 threat register, MITRE ATLAS mapping, L x I scoring, mitigations, residual risk, test cases

    OWASP Top 10 for Agentic Applications 2026; MITRE ATLAS (monthly, v2026.10)
Bước 4

Đánh giá và quản trị

  1. ISO/IEC 42001 Statement of Applicability Template (AIMS Annex A)

    Đầy đủ

    AIMS SoA: scope, risk treatment link, all 38 Annex A controls with basis, justification, status, owner, evidence, Annex B ref, extra controls, traceability

    ISO/IEC 42001:2023 clause 6.1.3, Annex A