Introduction & Scope
Welcome to GenFlow (“GenFlow”, “we”, “us”, or “our”), operated under the digital namespace usegenflow.io. GenFlow provides an enterprise-grade visual workflow automation and pipeline compiler platform tailored for creative engineering teams orchestrating multi-modal artificial intelligence models (such as Google Veo, Google Imagen, and Gemini via Google Cloud Vertex AI).
This Privacy Policy governs your use of GenFlow web interfaces, programmatic APIs, execution runners, and integrated cloud runtime environments. It describes our protocols regarding the collection, processing, transfer, storage, and cryptographical deletion of technical metadata, credentials, and customer assets.
By accessing or utilizing our workflow canvas, playground, orchestration nodes, or API endpoints, you acknowledge the terms established herein. For enterprise customers operating under a signed Master Services Agreement (MSA), our customized Data Processing Addendum (DPA) shall govern where explicitly conflicting.
Information We Collect
We adhere strictly to data minimization standards. The types of data processed through GenFlow are cataloged according to the following operational classifications:
Identity & Account Data
Names, verified business email addresses, company identity, federated SSO tokens (Google Workspace, Okta, SAML 2.0), cryptographic authentication keys, and role assignments.
Pipeline & Node Manifests
Workflow schematics, execution graph topology, parameter configurations (aspect ratio, CFG scale, seed tokens, model iterations), and automation triggers.
Customer Input & Media Payloads
Text prompts, translation script strings, character anchor reference imagery, UGC short-form source video files, and downstream generated raster/vector outputs.
Execution Telemetry
Node runtime durations, GPU compute quota consumed, API invocation latency, error status codes, IP addresses, browser agent headers, and audit ledger logs.
AI Models & Vertex AI Data Flow
Vertex AI Enterprise SLAThe GenFlow Zero-Training Guarantee
We execute model inference through dedicated, enterprise-tier Google Cloud Vertex AI instances. As part of Google Cloud’s contractual enterprise privacy guarantees and GenFlow’s orchestrator safeguards:
- No Foundational Model Retraining: Neither Google nor GenFlow utilizes your prompts, raw video reference plates, brand scripts, character models, or output assets to train, retrain, or align foundation models (including Veo 2.0/3.1, Imagen 4.0, or Gemini).
- Stateless Model Pipelines: Generative inference passes through ephemeral computational containers. Ingested weights and tokens are flushed immediately after pipeline execution completes.
- Regional Data Residency: Enterprise tenants may pin inference and storage endpoints to defined geographic regions (e.g., US-Central, EU-West, APAC-South) to maintain compliance with sovereign data mandates.
RBAC Sandboxed
Ephemeral Tokenizer
Zero Foundational Training
All communications between GenFlow dispatch proxies and Vertex AI runtime clusters leverage mTLS authentication and hardware security modules (HSMs).
How We Utilize Customer Data
GenFlow utilizes collected and ingested assets strictly to deliver, monitor, and invoice our generative automation services:
- Execution of Directed Workflows: Routing prompts, references, character poses, and upscaling configurations into Google Cloud Vertex AI instances per your pipeline definition.
- Multi-User Collaboration & Library Persistence: Enabling team members within your organization to catalog workflow recipes, view shared asset libraries, review output versions, and trace generation parameters.
- Security, Rate Limiting & Anomaly Detection: Detecting volumetric denial-of-service attempts, mitigating automated credential stuffing, and enforcing organizational API rate envelopes.
- Metering & Billing Calculation: Tracking generation seconds, model token usage, and upscaler invocations to generate transparent, auditable billing invoices.
Multi-Tenant Architecture & RBAC
GenFlow was engineered from inception to serve multi-organization enterprises. Data segregation is enforced at multiple architectural layers:
Logical Workspace Boundary
Each organization is mapped to an isolated row-level tenant identifier enforced by cryptographically signed database session variables. Zero cross-workspace leaking occurs at any database abstraction layer.
Role-Based Access Control (RBAC)
Fine-grained permission matrices determine whether an operator can author workflows, run production batches, inspect billing records, or export raw media assets from the asset archive.
Data Retention & Permanent Deletion
Enterprise administrators retain complete control over asset lifecycles. We enforce configurable retention schedules and provide rigorous cryptographic wiping upon contract completion:
| Data Classification | Default Retention | Customizable | Purge Method |
|---|---|---|---|
| Intermediate Frame Caches | 72 Hours | Yes (0–72h) | Cryptographic Shredding |
| Generated Asset Libraries | Duration of Subscription | Yes | Cloud Storage Hard Delete |
| Pipeline Execution Logs | 90 Days | Yes (Enterprise SLA) | Automated TTL Purge |
| User Authentication Credentials | Duration of Account + 30 Days | No | Database Tombstoning |
Subprocessors & Cloud Infrastructure
To deliver highly reliable, sub-second generative pipeline executions, GenFlow engages with vetted sub-processors meeting international data protection benchmarks:
Enterprise Security & Compliance Standards
Security is not bolted onto GenFlow; it is the foundation of our execution engine. We maintain a defense-in-depth posture:
Data In Transit
Enforced TLS 1.3 protocol with HTTP Strict Transport Security (HSTS) across all API routes and dashboard sessions.
Data At Rest
AES-256 bit encryption using customer-managed encryption keys (CMEK) option or Google Cloud KMS managed keys.
Audit Trail Logging
Immutable, tamper-evident security audit logs recording every login, prompt execution, API dispatch, and role adjustment.
Annual Third-Party Audits
Regular independent penetration testing and ongoing continuous compliance monitoring for SOC 2 Type II assurance.
Your Jurisdictional Privacy Rights
Depending on your physical operating jurisdiction (including the European Economic Area, UK, and California), you possess distinct statutory rights regarding personal information:
- Right of Access & Portability: You may request an export of your account data, workflows, and generated media library in standard machine-readable JSON/ZIP formats.
- Right to Erasure (“Right to be Forgotten”): You may request the permanent elimination of your team identity records and staging video caches.
- Right to Restrict or Object: You may object to specific automated processing pipelines or revoke previously granted consent credentials.
Contact & Data Protection Officer
For questions regarding this Privacy Policy, submitting Data Subject Access Requests (DSARs), or executing an Enterprise Data Processing Agreement (DPA), please reach our dedicated data compliance squad: