AI transparency

How Sage AI works

Sage runs on Anthropic Claude for generation and OpenAI for embeddings. No customer data is used to train models, and every AI action is metered and visible to you.

Explore the drafting and analysis support available through Sage.

See the advisory agentsExplore the platform

Four things you can hold us to

  • Disclosed providers

    Generation runs on Anthropic Claude; search embeddings run on OpenAI. Both providers are named.

  • No training on your data

    No customer data is used to train AI models.

  • Metered and visible

    Every AI action is metered and visible to you.

  • Domain-specific assistance

    Advisory agents support domain-specific GRC work. Their advisory scope is listed below.

Domain advisory agents

AI output is a draft for human review. Availability depends on your organisation’s configuration.

  • Evidence agent

    Suggest evidence classifications from the details you provide.

  • Policy agent

    Draft policies and procedures for review.

  • Questionnaire agent

    Draft suggested answers to security questionnaire questions.

  • Risk agent

    Suggest risk classifications from the details you provide.

  • Control agent

    Draft control-to-framework mapping suggestions.

  • Audit agent

    Draft audit plans from your scope and requirements.

  • Vendor risk agent

    Suggest vendor tiers from the details you provide.

  • Incident response agent

    Suggest incident classifications from the details you provide.

  • Regulatory intelligence agent

    Draft change summaries from regulatory material you provide.

  • Privacy agent

    Draft content for data subject access responses.

  • Training agent

    Draft training material from your topic and criteria.

  • Learning agent

    Draft study paths for your learning goals.

  • Insight agent

    Draft executive summaries from the information you provide.

  • ESG agent

    Draft ESG guidance for review.

  • SOX/ICFR agent

    Draft SOX and internal-control guidance for review.

  • AI governance agent

    Draft AI-governance guidance for review.

  • CISO Copilot

    Draft board-pack analysis for security leaders.

  • DPO Advisor

    Draft data protection impact assessment templates.

  • DORA agent

    Draft DORA advisory text for review.

Published security posture

  • Per-tenant isolation enforced in the database with row-level security
  • Encrypted sensitive fields
  • Audit logs record actions in the workspace
  • Data hosted in India
Read the security page

Frequently asked questions

Which AI providers power Sage AI?

Generation runs on Anthropic Claude and search embeddings run on OpenAI. Both providers are named.

Is my data used to train AI models?

No. No customer data is used to train models.

How is AI usage tracked?

Every AI action is metered and visible to you, and audit logs record actions in the workspace.

What do the domain advisory agents cover?

Sage includes advisory agents for Evidence, Policy, Questionnaire, Risk, Control, Audit, Vendor risk, Incident response, Regulatory intelligence, Privacy, Training, Learning, Insight, ESG, SOX/ICFR, AI governance, CISO Copilot and DPO Advisor work, plus DORA. Each produces a suggestion or draft.

Where is my data hosted?

Data is hosted in India. Per-tenant isolation is enforced in the database with row-level security.