Where do you stand? › Route 3 · Vendors & stack
I know which systems I need — who supplies them?
Pick the components you have to put in place. For each one you get the build-vs-buy reading and the market layer that supplies it, with the same scored recommendations and confidence the full analysis uses. Nothing is stored; the selection lives in the URL.
Target market(s)European UnionUnited States (federal)change
Legally-driven components are flagged when their requiring regulation sits outside your selected markets.
density
Step 1 of 2 — pick your components4 selected
Output Rails / Groundedness CheckHITL Escalation Queue & Review UIWORM / Immutable Audit VaultInterface Transparency & Content-Marking Layer
Step 2 of 2 — the vendor & stack view
4 of 4 selected components are covered by 5 market layers.
Named vendors are community-maintained, disputable examples — not an endorsement. The stable object is the market layer. Compare with the reference stack for your regulatory profile →
Build or buy, per component (4)
HITL Escalation Queue & Review UI buy (products exist) KRCA
A Agent Orchestration & SDLC Toolkits product can carry this; the buyer's duties stay with you.
Required by: AI Framework Act (KR), Quebec Law 25 (CA) in forcein force
Legally required in KR, CA — none of your selected markets carry this driver; treat it as a market-specific requirement, not a universal one. Help verify coverage →
Interface Transparency & Content-Marking Layer buy (products exist) CNIN
A Runtime Security & Guardrail Vendors / Public Transparency Registers & System Cards product can carry this; the buyer's duties stay with you.
Required by: AI-Generated Synthetic Content Labeling Measures (CN), IT Act 2000 + Intermediary Guidelines 2021 (IN) in forcein force
Legally required in CN, IN — none of your selected markets carry this driver; treat it as a market-specific requirement, not a universal one. Help verify coverage →
Output Rails / Groundedness Check buy (products exist) GBCN
A Runtime Security & Guardrail Vendors / Runtime Guardrails & Enforcement product can carry this; the buyer's duties stay with you.
Required by: Crime and Policing Act 2026 (GB), Interim Measures for Generative AI Services (CN), Online Safety Act 2023 (GB) unverified — help verifyin forcein force
Legally required in GB, CN — none of your selected markets carry this driver; treat it as a market-specific requirement, not a universal one. Help verify coverage →
WORM / Immutable Audit Vault buy (products exist) EUDEthinUS
A Cryptographic Evidence & Audit Ledger product can carry this; the buyer's duties stay with you.
Required by: AI Liability Directive (withdrawn), BaFin MaRisk (Mindestanforderungen an das Risikomanagement), Bank Secrecy Act / FinCEN Program Rules, FINRA Rule 4511 (General Books & Records), Revised Product Liability Directive, SEC Rule 17a-4 (US Records Retention) withdrawn — no longer lawin forcein forcein forcein forcein force
Legally required in EU, US (your selected markets); also required in DE, which you have not selected. Help verify coverage →
Runtime Security & Guardrail Vendors — covers 2 of your components
Covers: Output Rails / Groundedness Check, Interface Transparency & Content-Marking Layer. This layer supplies 7 components in the graph.
Confidence: strong (81/100)
Community-maintained examples
Lakera · HiddenLayer · Palo Alto Prisma AIRS · Cisco AI Defense · NVIDIA NeMo Guardrails · Guardrails AI · Garak · Protect AI
Filters to self-hostable, customer-VPC and open-source options when personal or confidential data cannot leave the EU.
| Example | Sub-category | What it does | Hosting | Claimed alignments |
|---|---|---|---|---|
| Lakera | guardrail proxy | Inline prompt-injection and content detection at request time. Typical: injection defence, content filtering. | not checked | SOC 2 (claimed)supports Art. 15 robustness measures (claimed) |
| HiddenLayer | model/agent detection & response | Model-layer detection and response with adversarial-attack telemetry. Typical: model threat detection, red-team telemetry. | not checked | SOC 2 (claimed)supports Art. 15 robustness measures (claimed) |
| Palo Alto Prisma AIRS | network-integrated AI security | AI runtime security folded into an existing enterprise network security estate. Typical: enterprise rollout, egress control. | not checked | SOC 2 (claimed)enterprise security integration (claimed) |
| Cisco AI Defense | network-integrated AI security | Discovery of AI usage plus inline enforcement across the corporate network. Typical: shadow-AI discovery, inline enforcement. | not checked | enterprise security integration (claimed) |
| NVIDIA NeMo Guardrails | open guardrail framework | Programmable dialogue and action rails, self-hostable alongside your models. Typical: dialogue rails, action gating. | open source | supports Art. 15 robustness measures (claimed) |
| Guardrails AI | open guardrail framework | Open validator library for structured output checks and policy validators. Typical: output validation, schema enforcement. | open source | OSS, no vendor certification |
| Garak | adversarial scanner | Open-source LLM vulnerability scanner used for pre-deployment probing. Typical: red-teaming, release gating. | not checked | OSS, no vendor certificationsupports Art. 15 testing evidence (claimed) |
| Protect AI | ML supply-chain & model security | Model scanning and ML supply-chain security tooling (Palo Alto Networks acquisition reported 2025). Typical: model scanning, supply-chain security. | SaaS (vendor cloud) | supports Art. 15 cybersecurity measures (claimed) |
Community-maintained, disputable examples — not an endorsement and not a ranking. Alignments are as claimed by vendors or the source compilation, not verified by RAIN; a certification is shown as a certification only where a certificate or registry reference is recorded.
Disclosure: RAI·N·avigator operates in this category too, so we have a commercial interest in any comparison here. That is why this layer maps product classes to control objectives and lists named products as community-maintained examples — we publish no rankings, no quadrants and no coverage assertions about any vendor, including ourselves.
Select on
single-pass parallel evaluation latency (<20 ms class), injection/hallucination catch rates, SecOps/SIEM routing, policy versioning surfaced into the AI-BOM
Why this confidence
- 2 in-scope components of this use case are supplied by this layer (Output Rails / Groundedness Check, Interface Transparency & Content-Marking Layer) — a direct supplied_by path in the graph.
- The catalog use-case match is strong, so the component set this layer was derived from is reliable.
- High-risk tier: this layer carries mandatory Chapter III duties, so some tooling in it is non-optional.
- 8 community-maintained example vendors recorded on the layer node.
- Selection metrics for this layer are documented, so the shortlist can be compared objectively.
- Only 2 of 7 components this layer supplies are in your scope — evaluate a narrow subset of its capabilities.
Alternatives
- Public Transparency Registers & System Cards — confidence strong (70/100, -11 vs. this layer); overlapping coverage, no additional selected component
- Cryptographic Evidence & Audit Ledger — confidence moderate (68/100, -13 vs. this layer); also covers: WORM / Immutable Audit Vault
Public Transparency Registers & System Cards — covers 1 of your components
Covers: Interface Transparency & Content-Marking Layer. This layer supplies 2 components in the graph.
Confidence: strong (70/100)
Community-maintained examples
Saidot
Filters to self-hostable, customer-VPC and open-source options when personal or confidential data cannot leave the EU.
| Example | Sub-category | What it does | Hosting | Claimed alignments |
|---|---|---|---|---|
| Saidot | public AI register | AI register with published system cards and regulation-mapped documentation workflows. Typical: public AI register, system cards. Scope overlap: Its documentation and register scope overlaps this platform's own; we have a commercial interest in the comparison. | SaaS (vendor cloud) | EU AI Act documentation positioningISO 42001 alignment (claimed) |
Community-maintained, disputable examples — not an endorsement and not a ranking. Alignments are as claimed by vendors or the source compilation, not verified by RAIN; a certification is shown as a certification only where a certificate or registry reference is recorded.
Disclosure: RAI·N·avigator operates in this category too, so we have a commercial interest in any comparison here. That is why this layer maps product classes to control objectives and lists named products as community-maintained examples — we publish no rankings, no quadrants and no coverage assertions about any vendor, including ourselves.
Select on
Versioning of published statements against the system version they describe; whether a card is generated from your governance record or re-authored by hand; language coverage and accessibility of the published surface; export and self-hosting of the public register; whether unpublishing leaves an auditable trail.
Why this confidence
- 1 in-scope component of this use case is supplied by this layer (Interface Transparency & Content-Marking Layer) — a direct supplied_by path in the graph.
- This layer's graph purpose overlaps strongly with your scope (1 of 2 components it supplies are in scope).
- The catalog use-case match is strong, so the component set this layer was derived from is reliable.
- High-risk tier: this layer carries mandatory Chapter III duties, so some tooling in it is non-optional.
- 1 community-maintained example vendor recorded on the layer node.
- Selection metrics for this layer are documented, so the shortlist can be compared objectively.
Alternatives
- Runtime Security & Guardrail Vendors — confidence strong (81/100, +11 vs. this layer); also covers: Output Rails / Groundedness Check
- Cryptographic Evidence & Audit Ledger — confidence moderate (68/100, -2 vs. this layer); also covers: WORM / Immutable Audit Vault
Cryptographic Evidence & Audit Ledger — covers 1 of your components
Covers: WORM / Immutable Audit Vault. This layer supplies 3 components in the graph.
Confidence: moderate (68/100)
Community-maintained examples
Fact0 · Traccia
Filters to self-hostable, customer-VPC and open-source options when personal or confidential data cannot leave the EU.
| Example | Sub-category | What it does | Hosting | Claimed alignments |
|---|---|---|---|---|
| Fact0 | cryptographic evidence ledger | Positions itself as a tamper-evident ledger for AI decision records. Typical: decision records, audit trail. | not checked | supports Art. 12 record-keeping (claimed) |
| Traccia | audit trail & traceability | Positions itself around traceability of AI pipeline steps and artefacts. Typical: traceability, artifact lineage. | not checked | supports Art. 12 record-keeping (claimed) |
Community-maintained, disputable examples — not an endorsement and not a ranking. Alignments are as claimed by vendors or the source compilation, not verified by RAIN; a certification is shown as a certification only where a certificate or registry reference is recorded.
Disclosure: RAI·N·avigator operates in this category too, so we have a commercial interest in any comparison here. That is why this layer maps product classes to control objectives and lists named products as community-maintained examples — we publish no rankings, no quadrants and no coverage assertions about any vendor, including ourselves.
Select on
Append-only guarantees and who can rotate or delete (including the vendor); anchoring mechanism (qualified timestamp, transparency log, notarisation) and whether verification works without the vendor; retention and export in a readable format at end of contract; throughput and cost at your event volume.
Why this confidence
- 1 in-scope component of this use case is supplied by this layer (WORM / Immutable Audit Vault) — a direct supplied_by path in the graph.
- The catalog use-case match is strong, so the component set this layer was derived from is reliable.
- High-risk tier: this layer carries mandatory Chapter III duties, so some tooling in it is non-optional.
- 2 community-maintained example vendors recorded on the layer node.
- Selection metrics for this layer are documented, so the shortlist can be compared objectively.
- Only 1 of 3 components this layer supplies are in your scope — evaluate a narrow subset of its capabilities.
Alternatives
- Runtime Security & Guardrail Vendors — confidence strong (81/100, +13 vs. this layer); also covers: Output Rails / Groundedness Check, Interface Transparency & Content-Marking Layer
- Public Transparency Registers & System Cards — confidence strong (70/100, +2 vs. this layer); also covers: Interface Transparency & Content-Marking Layer
Agent Orchestration & SDLC Toolkits — covers 1 of your components
Covers: HITL Escalation Queue & Review UI. This layer supplies 4 components in the graph.
Confidence: moderate (67/100)
Community-maintained examples
LangChain / LangGraph · LlamaIndex · Microsoft AutoGen · CrewAI · DSPy · Semantic Kernel · PydanticAI · Model Context Protocol (MCP) · E2B · Airia
Filters to self-hostable, customer-VPC and open-source options when personal or confidential data cannot leave the EU.
| Example | Sub-category | What it does | Hosting | Claimed alignments |
|---|---|---|---|---|
| LangChain / LangGraph | agent framework | Graph-structured agent runtime; interrupt/pause nodes support implementing human approval at defined steps. Typical: multi-step agents, approval workflows. | not checked | supports implementing Art. 14 oversight (claimed)supports Art. 12 step logging (claimed) |
| LlamaIndex | RAG framework | Indexing and query abstractions over documents and structured sources. Typical: enterprise RAG, document agents. | open source | retrieval-governance positioning |
| Microsoft AutoGen | multi-agent framework | Conversational multi-agent patterns with pluggable tool executors. Typical: multi-agent research, code agents. | not checked | research/OSS, no vendor certification |
| CrewAI | multi-agent framework | Role-based agent teams with task delegation and process templates. Typical: process automation, role-based agents. | not checked | vendor-stated security posture |
| DSPy | prompt/program optimisation | Declarative programs with optimisers that make prompt changes reproducible and testable. Typical: evaluated pipelines, model validation. | not checked | model-validation positioning (SR 11-7 style, claimed) |
| Semantic Kernel | enterprise SDK | Microsoft SDK for planners and plugins inside .NET/Java estates. Typical: enterprise copilots, tool plugins. | not checked | enterprise-estate integration (claimed) |
| PydanticAI | typed agent SDK | Type-validated agent outputs and tool signatures for deterministic contracts. Typical: structured outputs, typed tool calls. | not checked | schema-enforcement positioning |
| Model Context Protocol (MCP) | protocol / standard | Open protocol for tool and context exposure; a protocol, not a product — governance sits in the gateway around it. Typical: tool interoperability, gateway mediation. | not checked | open protocol, no certification |
| E2B | sandboxed runtime | Ephemeral cloud sandboxes for agent code execution with isolation per task. Typical: code agents, untrusted execution. | not checked | isolation/sandbox positioning |
| Airia | enterprise agent platform | Enterprise platform for building and running agents with connector, policy and routing layers. Typical: agent orchestration, internal copilots. | SaaS (vendor cloud) | EU AI Act readiness positioningSOC 2 programme positioning (claimed) |
Community-maintained, disputable examples — not an endorsement and not a ranking. Alignments are as claimed by vendors or the source compilation, not verified by RAIN; a certification is shown as a certification only where a certificate or registry reference is recorded.
Disclosure: RAI·N·avigator operates in this category too, so we have a commercial interest in any comparison here. That is why this layer maps product classes to control objectives and lists named products as community-maintained examples — we publish no rankings, no quadrants and no coverage assertions about any vendor, including ourselves.
Select on
broad model-API abstraction, state/memory management, error recovery, fallback routing hooks
Why this confidence
- 1 in-scope component of this use case is supplied by this layer (HITL Escalation Queue & Review UI) — a direct supplied_by path in the graph.
- The catalog use-case match is strong, so the component set this layer was derived from is reliable.
- High-risk tier: this layer carries mandatory Chapter III duties, so some tooling in it is non-optional.
- 10 community-maintained example vendors recorded on the layer node.
- Selection metrics for this layer are documented, so the shortlist can be compared objectively.
- Only 1 of 4 components this layer supplies are in your scope — evaluate a narrow subset of its capabilities.
Alternatives
- Runtime Security & Guardrail Vendors — confidence strong (81/100, +14 vs. this layer); also covers: Output Rails / Groundedness Check, Interface Transparency & Content-Marking Layer
- Public Transparency Registers & System Cards — confidence strong (70/100, +3 vs. this layer); also covers: Interface Transparency & Content-Marking Layer
Runtime Guardrails & Enforcement — covers 1 of your components
Covers: Output Rails / Groundedness Check. This layer supplies 4 components in the graph.
Confidence: moderate (67/100)
Community-maintained examples
Guardrails AI · NVIDIA NeMo Guardrails · Lakera AI · Credal AI
Filters to self-hostable, customer-VPC and open-source options when personal or confidential data cannot leave the EU.
| Example | Sub-category | What it does | Hosting | Claimed alignments |
|---|---|---|---|---|
| Guardrails AI | validation framework | Open-source validator framework for structured output and content policies in the request path. Typical: output validation, structured output. | open source | supports Art. 15 robustness measures (claimed) |
| NVIDIA NeMo Guardrails | dialogue policy rails | Programmable dialogue and topic rails placed around an LLM application. Typical: topic control, dialogue policy. | open source | supports Art. 50 interaction disclosure patterns (claimed) |
| Lakera AI | guardrail proxy | Inline prompt-injection and content detection at request time. Typical: injection defence, content filtering. | SaaS (vendor cloud) | SOC 2 (claimed)supports Art. 15 robustness measures (claimed) |
| Credal AI | enterprise access & policy layer | Permission-aware access layer with data-loss controls in front of enterprise assistants. Typical: access control, DLP. | SaaS (vendor cloud) | SOC 2 (claimed) |
Community-maintained, disputable examples — not an endorsement and not a ranking. Alignments are as claimed by vendors or the source compilation, not verified by RAIN; a certification is shown as a certification only where a certificate or registry reference is recorded.
Disclosure: RAI·N·avigator operates in this category too, so we have a commercial interest in any comparison here. That is why this layer maps product classes to control objectives and lists named products as community-maintained examples — we publish no rankings, no quadrants and no coverage assertions about any vendor, including ourselves.
Select on
Where enforcement sits (inline proxy, sidecar, SDK) and the added latency at your token volumes; whether policy is versioned and testable as code; fail-open vs. fail-closed behaviour under guardrail outage; language and modality coverage; whether every block writes an evidence record you can cite later.
Why this confidence
- 1 in-scope component of this use case is supplied by this layer (Output Rails / Groundedness Check) — a direct supplied_by path in the graph.
- The catalog use-case match is strong, so the component set this layer was derived from is reliable.
- High-risk tier: this layer carries mandatory Chapter III duties, so some tooling in it is non-optional.
- 4 community-maintained example vendors recorded on the layer node.
- Selection metrics for this layer are documented, so the shortlist can be compared objectively.
- Only 1 of 4 components this layer supplies are in your scope — evaluate a narrow subset of its capabilities.
Alternatives
- Runtime Security & Guardrail Vendors — confidence strong (81/100, +14 vs. this layer); also covers: Interface Transparency & Content-Marking Layer
- Public Transparency Registers & System Cards — confidence strong (70/100, +3 vs. this layer); also covers: Interface Transparency & Content-Marking Layer