Skip to content

Trust center

This is the consolidated trust summary for Venturi: the single page a prospective customer’s security and procurement reviewer can read to get the answer to every common diligence question, with a pointer to the detailed page behind each answer. Every row here is defensible from the rest of this section.

Rows describe Venturi’s controls and attestation status. Tenant-specific configuration and verification evidence, and separately identified roadmap items, are explained in Docs authority & product state.

Trust summary

Trust question Venturi’s answer
Where does our data live? In a dedicated single-tenant data plane: your cloud account in self-hosted mode, or the contracted residency lane in SaaS mode. → Security architecture
Can Venturi reach into our network? No. The dedicated data plane initiates the narrow, signed connection to the region-pinned control plane; the control plane initiates nothing inbound. Full offline mode is available. → Security architecture
Can Venturi change anything in our systems? No. Connectors only read: they request read-only scopes wherever the provider offers them, and provider-account connectors send only GET requests to allowlisted read endpoints with a dedicated credential and its narrowest available permissions. OpenAI organization Admin keys support read-permission selection; Anthropic reporting credentials can carry wider management capability, which the connector’s independent egress boundary must exclude. The IAM policies on Venturi’s workloads carry an explicit Deny on write actions against your source systems and cloud resources, the customer cross-account role grants no write permission, and no flag grants write access. → Security architecture
Can Venturi take our AI traffic down? A failing Venturi dependency cannot. The decision-time interceptor is fail-open within a 50 ms P99 end-to-end budget, bounded by a cooperative timeout against a cumulative deadline: if the index lookup, the event stream, or any other Venturi dependency behind it is slow, degraded, or unavailable, it forwards your request unmodified. Fail-open is not configurable. The serving-plane SLO (99.9% over a rolling 28-day window) bounds attribution availability, not the forwarding of your AI traffic. With the drop-in proxy, the proxy endpoint itself is on your request path, and no availability commitment covers it; configure client-side failover to the provider’s direct base URL, or capture through the Ingestion API or a connector. → Security architecture and Proxy endpoint reachability
Do you store our prompts/completions? No. Some provider log sources can contain prompts and completions: Amazon Bedrock invocation logs, Vertex AI request-response logs, and Azure API Management language-model logs when you enable them, and Vertex AI container logs (on by default for v1 endpoints) whenever your serving container writes content to standard output or standard error. The collector discards content fields in memory before a record is materialized and never persists them, and the normalized schema contains metadata only. On a call routed through the drop-in proxy, request and response content exists only in memory while it is forwarded and is never written to logs, traces, queues, or exception payloads. Provider API keys on proxied calls are never stored: Venturi keeps only a one-way keyed fingerprint and a truncated display prefix. → Content-bearing logs
Do you profile or score individual employees? No. Adoption intelligence is cohort-only: every cohort read path enforces a server-side minimum cohort of 5, and individual-level views do not exist. No emotion or behavioral-state inference. → Data privacy and retention
How is our data isolated from other tenants? Per-tenant database, event topics, and a dedicated KMS key per tenant; cross-tenant queries are rejected with 403 TENANT_MISMATCH and logged. These are activation requirements. → Tenant isolation
Can Venturi support staff see our data? Not by default. There is no standing support access. Break-glass access requires a customer Admin’s approval for each request and is time-boxed (4 hours by default, 24 hours at most), MFA-gated, read-only by default, audit-logged, and notified to you. → Tenant isolation
How is data encrypted? AES-256 at rest with your KMS key; TLS 1.3 in transit; mTLS (SPIFFE) for internal gRPC; certificate-pinned mTLS on the data-plane-initiated control-plane connection. → Security architecture
How are secrets managed? A managed secrets store; never on disk or in pod env; database credentials rotate every 90 days; JWT signing keys are externally managed and rotated; your provider keys never reach the Venturi control plane in plaintext. Cloud onboarding uses no long-lived cloud credentials by default; the Azure client-secret fallback, Venturi API keys, and SaaS connector credentials are the named exceptions. → Security architecture
What about audit logs? An append-only, pseudonymized audit log records every policy decision, model promotion, override, and break-glass access. It is stored under S3 Object Lock with a 5-year retention period: Governance mode by default, or Compliance mode, configured at onboarding where required, in which no user (including the AWS account root user) can delete a locked entry before it expires. → Data privacy & retention
How long do you keep our data, and can we delete it? Operational retention defaults to 13 months (configurable from 30 days to 5 years). Erasure is by crypto-shredding your per-subject or per-tenant data-encryption keys within 30 days, with a deletion certificate. → Data privacy & retention
How do you handle a right-to-erasure request? On your instruction, we erase operational and attribution data by crypto-shredding within 30 days. Audit and override stores remain pseudonymized under a named lawful basis. Anonymized aggregation contributions are non-personal and cannot be reversed individually. → Erasure details
Are you SOC 2 certified? Readiness program in progress; report not yet issued. SOC 2 Type I readiness/report target on our security roadmap with Type II targeted within 12 months; it is not yet held. Venturi does not hold a SOC 2 report or attestation. The scope is Security, Availability, Confidentiality, and Processing Integrity; Privacy is added to the audit scope for EU or regulated-sector customer engagements. → Compliance
Do you run penetration tests? Yes. A qualified third party tests the API endpoints, authentication flows, and tenant data boundaries at least annually, and critical and high findings are remediated before the next customer deployment. The executive summary and remediation status are in the Trust Center and the onboarding diligence pack; the full report is available on request under NDA. → Penetration testing
What is your GDPR role? Venturi is your processor; you are the controller. The Data Processing Agreement (DPA) carries the Art. 28(3) terms and the Art. 28(2) subprocessor change notice with an objection window. The onboarding diligence pack adds the RoPA summary, a DPIA template and works-council pack, and a lawful-basis mapping. → Compliance
GDPR and data residency? Data stays in the dedicated tenant data plane and contracted residency lane; applicable transfers use Chapter V safeguards. → Residency and subprocessors
EU AI Act? Deployment-specific assessment determines the applicable provider/deployer roles, classification, obligations, and effective dates under the law in force. Venturi completes the applicable assessment and market-entry obligations; cohort-only adoption intelligence and no emotion inference are product controls, not a legal classification. → Compliance
DPA and subprocessors? The onboarding diligence pack includes the DPA, the DPIA template and works-council pack, and the dated, versioned subprocessor list for your deployment. The in-product Trust Center shows the current list and every earlier version. → Residency and subprocessors
Incident response & breach notification? A documented Sev-1 to Sev-4 incident plan: each severity sets a first-notice deadline and an update cadence, from a first notice within 1 hour and an update every hour for Sev-1 to a first notice within 24 hours and an update every 24 hours for Sev-4. Venturi notifies you of a personal-data breach without undue delay (target ≤ 24 h of becoming aware) so you can meet your GDPR Art. 33 72-hour deadline. Operational incidents appear on the tenant-isolated in-product status surface; there is no shared public status page. → Compliance
How is the platform threat-modeled? A STRIDE threat model maps each threat to its control and residual risk, and a red-team exercise tests it at least annually. → Security architecture
Can we trust the numbers for chargeback? Operational confidence (coper) has an auditable policy ceiling of 0.95 and is never raised. The Chargeback-ready band (coper >= 0.80) is a necessary confidence floor; export_eligibility, qualifying output state, current evidence, no open dispute or degradation, reconciliation, and finance acceptance also govern chargeback. Provisional attributions (0.50 <= coper < 0.80) are labeled and never auto-billed; unknown is reported as unknown, never zero. → Confidence and evidence
What about your supply chain? Dependency and image scanning with critical-finding remediation within 7 days; signed images and an SBOM per release; model artifacts are hash- and signature-verified at load. → Security architecture
How do we report a vulnerability? Email security@venturi.systems. Venturi operates a coordinated vulnerability disclosure program with safe-harbor terms for good-faith research. → Vulnerability disclosure policy

Control framework crosswalk

Venturi’s onboarding-baseline controls map to the principal governance frameworks. Venturi does not hold ISO/IEC 27001 (ISMS) or ISO/IEC 42001 (AIMS) certification, a SOC 2 report, or a FedRAMP authorization. The mapping supports your diligence against those frameworks and is not a certification claim.

The ISO/IEC column covers ISO/IEC 27001 and ISO/IEC 42001. Fairness by subgroup means subgroup-disaggregated evaluation. Human review covers model selection and disputes.

Venturi control NIST AI RMF NIST SP 800-53 / CSF ISO/IEC
Fail-open + per-plane SLOs + error budget MANAGE 2.x; MEASURE 2.x CP-10, SC-5; CSF RC.RP 27001 A.5.29, A.8.14; 42001
Confidence cap / calibration / false-high-confidence gate MEASURE 2.x SA-11; RA-3 42001
Fairness by subgroup MEASURE 2.11 N/A 42001
Read-only IAM deny / least privilege MANAGE 1.x AC-3, AC-6 27001 A.5.15, A.8.2
Tenant isolation / cross-tenant 403 MAP 5.x SC-2, SC-4, SC-7 27001 A.8.20 to A.8.23
Encryption at rest/in transit, KMS lifecycle N/A SC-8, SC-12, SC-13, SC-28 27001 A.8.24
Manage and rotate secrets N/A IA-5, SC-12 27001 A.5.17, A.8.24
Audit trail (append-only, pseudonymized, S3 Object Lock retention) GOVERN 1.x; MANAGE 4.x AU-2, AU-9, AU-11 27001 A.8.15, A.5.28
Retention / crypto-shred erasure / DSAR MAP 4.x SI-12; MP-6 27001 A.5.34; 42001
Threat model / red-team MEASURE 1.x; MANAGE 1.x SA-11, SA-15, RA-3, CA-8 27001 A.8.8, A.8.29
Supply chain / SBOM / signing MAP 4.1 RA-5, SR family; SSDF SP 800-218 27001 A.8.8, A.5.21
Human review GOVERN 1.x; MANAGE 3.x N/A 42001
EU AI Act assessment GOVERN 1.x N/A 42001

Human oversight

A single human-oversight requirement governs which decisions require a human: production model selection is a human Venturi-leadership decision recorded with an evidence card (an automated approver is rejected), and customer disputes and overrides are human-adjudicated and tracked against the dispute SLA clock you configure for your business calendar. Both are recorded in the audit trail.

Reliability commitments

The serving plane carries explicit SLOs, distinct from the fail-open behavior of the customer traffic path:

Objective Target
Customer AI traffic behind the interceptor Forwarded unmodified when a Venturi dependency fails (fail-open by design); the drop-in proxy endpoint itself is not covered, see Proxy endpoint reachability
Serving plane (query and dashboard API) availability 99.9% over a rolling 28-day window on every support tier; 99.95% under the Premium and Regulated tiers
Recovery point / time objective RPO ≤ 15 min / RTO ≤ 1 h
Index freshness P99 ≤ 90 s objective; an SLA of at most 5 minutes under the Premium and Regulated tiers
Reconciliation latency ≤ 24 h
Breach notification to you Target ≤ 24 h of becoming aware

The support tiers (Standard, Premium, and Regulated) are defined in Support, and your order form states the tier you hold. SLAs & SLOs gives the measurement basis for each objective.

A Business Continuity Plan extends disaster recovery to key-person succession, cloud-provider or region outages, and critical-subprocessor failure, and is exercised at least annually.

Status surface

Operational status is tenant-isolated and in-product. The app’s /status route shows component status, active incidents, scheduled maintenance, and recent uptime for your deployment without disclosing one customer’s deployment state to another. There is no shared public status page. Incident notices also reach you in-app and by email, Slack, and Microsoft Teams according to your tenant notification preferences. See Status & incidents.

Diligence artifacts & vulnerability disclosure

This section opens with the vulnerability disclosure policy, then lists the diligence evidence that security and procurement reviewers receive under NDA.

Vulnerability disclosure policy

Venturi operates a coordinated vulnerability disclosure program. The RFC 9116 security.txt file names the current security contact, and its Policy field links to this section.

  • How to report. Email security@venturi.systems with the affected component or URL, the steps to reproduce, and the impact you observed.
  • What happens next. Venturi acknowledges the report, opens a tracked security case for it, keeps you informed through remediation, and coordinates the disclosure timing with you.
  • Safe harbor. Venturi treats good-faith research conducted under this policy as authorized and does not pursue legal action against it. Good-faith research avoids privacy violations, data destruction, and service degradation; uses only the access needed to demonstrate the issue; never accesses another customer’s tenant data; and gives Venturi reasonable time to remediate before public disclosure.

Venturi does not offer a paid bug bounty.

Onboarding diligence pack

Prospective customers receive the onboarding diligence pack under NDA before signing. Request it from your Venturi representative or at security@venturi.systems. The pack contains:

Artifact What it covers
Architecture and data-flow diagrams The trust boundary, the data-plane-initiated control-plane connection, and where each data class flows.
Control matrices and the control framework crosswalk Each control and its mapping to NIST AI RMF, NIST SP 800-53 / CSF, and ISO/IEC 27001 / 42001.
Threat model summary The STRIDE threats, their controls, and residual risk.
SOC 2 status statement SOC 2 Type I readiness/report target on our security roadmap with Type II targeted within 12 months; it is not yet held.
Penetration-test executive summary and remediation status The latest third-party test; the full report is available on request under NDA.
Data Processing Agreement (DPA) The Art. 28(3) terms, with Standard Contractual Clauses where GDPR Chapter V applies.
Record of Processing Activities (RoPA) summary Processing activities, purposes, data categories, lawful basis, and recipients.
DPIA template (GDPR Art. 35) and works-council pack Templates your privacy team and works council adapt to your jurisdiction.
Lawful-basis mapping The lawful basis for each processing activity.
Dated, versioned subprocessor list Each vendor, purpose, data category, processing region, and transfer safeguard for your deployment mode and residency lane.
Residency posture Your residency lane and the fail-closed residency routing.
Retention and deletion policy Retention defaults, crypto-shred erasure, and the offboarding sequence.
Business continuity and disaster recovery summary Recovery objectives and the annual continuity exercise.
Security questionnaire responses Completed, dated responses to the CSA CAIQ and Shared Assessments SIG questionnaires, each answer linked to its supporting control or evidence.
Information security policy index Each governed security policy with its version, owner, and review date, plus a redacted summary.
Third-party risk management program How Venturi tiers, reviews before onboarding, contractually binds, and periodically re-assesses its own vendors and subprocessors.
Vendor-risk profile Corporate and legal profile, financial-viability statement, and insurance attestation.

Onboarded tenants can also retrieve the following artifacts in the in-product Trust Center (Security & Compliance / Compliance / Trust Center), each at its current version with its effective date: the SOC 2 status statement, the penetration-test executive summary and remediation status, the architecture and data-flow diagrams, the DPA, the RoPA summary, the subprocessor list, the residency posture, the retention and deletion policy, the security questionnaire responses, the information security policy index, and the third-party risk management program. See Complete the trust-center review.

Penetration testing

A qualified third party performs a penetration test at least annually. Each test covers the API endpoints, authentication flows, and tenant data boundaries, and critical and high findings are remediated before the next customer deployment. The executive summary and remediation status of the latest test are in the onboarding diligence pack and the in-product Trust Center; the full report is available on request under NDA. See Compliance.

Give feedback on this page

Sign in to send feedback. Opens in a new tab.

Copy page details 
Docs feedback: Trust center
Page: https://docs.venturi.systems/security/trust-center/