Policy Version: 2026.1-draftDraft Preview

Privacy Policy

Effective Date: Not yet configured (Development Preview) • Last Updated: October 2026

Development Preview Draft Notice

This document describes actual, verified data handling practices implemented in the Rowforma software codebase. Production operator identification, registered business address, and dedicated privacy officer contact channels are not yet finalized and are marked as “Not yet configured”. This policy must be finalized and formally adopted before commercial or production launch.

Plain-Language Summary

Temporary CSV ProcessingCSVs uploaded for cleaning are stored in private temporary storage with an automatic 24-hour access expiry. Physical deletion occurs via periodic worker cleanup routines.
Zero AI Model TrainingRowforma is a deterministic rule and validation engine. Uploaded customer CSV data is never used to train external, public, or foundation AI models.
No Third-Party Ad TrackersWe do not sell personal data, do not use third-party marketing tracking pixels, and do not track users across unrelated websites.
Hashed Keys & Encrypted SecretsAPI keys are never stored in plaintext (only SHA-256 hashes). Customer webhook secrets are encrypted at rest using AES-256-GCM.

1. Scope and Roles

This Privacy Policy explains how Rowforma (“we”, “us”, or “our”), operated by Not yet configured (Development Preview), collects, uses, retains, and protects personal information and customer data when you access or use our web portal, developer APIs, documentation, and SDKs.

In the context of applicable Canadian and provincial data protection legislation, including the federal Personal Information Protection and Electronic Documents Act (PIPEDA) and the Alberta Personal Information Protection Act (PIPA):

  • For Account & Contact Information: Rowforma acts as the organization/controller determining the purposes and means of collecting your account credentials, billing contact details, and platform access logs.
  • For Uploaded CSV Datasets: You (or your organization) are the data controller. Rowforma processes uploaded tabular data strictly on your authorization and according to your instructions (the “service provider” or “data processor” role) to perform schema validation, issue remediation, and export generation.

2. Categories of Data Processed

We process only the minimum information necessary to provide deterministic CSV preparation services:

A. Account and Profile Identifiers

User email address, unique authentication identifier (Firebase UID), account/workspace membership, assigned role (e.g., owner, member), account creation timestamps, and record of Terms of Service acceptance version and timestamp. Passwords are handled exclusively by Firebase Authentication; Rowforma never sees or stores plaintext passwords.

B. Customer Uploaded Datasets

Raw tabular CSV files uploaded directly through the web portal, ingested via API endpoints, or submitted through the embedded importer component. This may include contact lists, inventory records, financial transactions, or customer-defined tabular data.

C. Analysis, Review Decisions, and Generated Artifacts

Structural metadata generated during file processing, including column names, detected data types, sample values, and validation violation summaries. In interactive review workflows, we store individual row/cell approval or rejection decisions and user-submitted cell edits. Upon completion, we generate downloadable output artifacts: cleaned ready.csv files and error-isolated unresolved.csv files.

D. Persistent Recipe and Schema Configurations

Column definitions, mapping targets, data validation rules (e.g., regex patterns, range limits, email formatters), transformation recipes, and starter template configurations established within your workspace.

E. Security Credentials and Webhook Metadata

API key identifiers, key names, and secure SHA-256 one-way cryptographic hashes. Webhook endpoint URLs, subscribed event topics, delivery attempt status codes, and customer-configured signing secrets encrypted at rest.

F. Operational Logs and Diagnostics

Server request logs including HTTP method, path, response status, duration, request identifiers, worker queue processing metrics, and error codes necessary for operational stability and denial-of-service mitigation.

3. Purposes and Lawful Grounds

We collect and process personal data exclusively for identified, legitimate business purposes:

  • Service Delivery: To authenticate your identity, establish isolated multi-tenant workspaces, parse and validate your uploaded files, execute customer-directed recipes, and deliver cleaned CSV artifacts.
  • Security and Abuse Prevention: To enforce API rate limits (fixed 1-minute window quotas), safeguard infrastructure against unauthorized access or volumetric denial-of-service, verify embed origin domains, and maintain audit logs.
  • Usage Accounting: To track monthly processed job volume and active concurrency slots against workspace allowances.
  • Communications: To deliver critical service notices, transaction responses, and password-reset messages initiated by you.

Notice Regarding Artificial Intelligence:

Rowforma does not use customer CSV data, parsed records, or remediation decisions to train, fine-tune, or calibrate public or foundation machine learning or artificial intelligence models.

4. Temporary File Storage & Cleanup

Rowforma is architected around short-term data lifecycles. We do not provide long-term cloud document archiving:

  • 24-Hour Access Expiry: Uploaded source CSV files and generated output artifacts (ready.csv, unresolved.csv) are assigned a strict 24-hour time-to-live from the moment upload commits. Upon expiry, all API endpoints and web interface links immediately refuse read access with HTTP 410 (Gone) or HTTP 404 (Not Found).
  • Configured Expiry vs. Physical Cleanup: While user-facing access terminates promptly at the 24-hour mark, physical file unlinking and deletion from underlying storage volumes is executed asynchronously by background worker routines during scheduled sweeps. We make no representations of exact-second physical sector overwriting.
  • User-Initiated Cancellation: When a user cancels a job before completion, associated temporary files are scheduled for immediate cleanup.

5. Persistent Metadata vs. Temporary Data

It is important to distinguish temporary file content from persistent platform metadata:

What Persists After File Expiry:

  • High-level job metadata: Job ID, recipe identifier, file byte size, total row count, processing duration, and aggregate record counts (e.g., number of valid, approved, rejected, and unresolved rows).
  • Workspace configuration: Saved recipes, published versions, field validation rules, and team member accounts.
  • Security records: API key identifiers, webhook endpoint definitions, delivery logs, and Terms acceptance timestamps.

Persistent job metadata is retained to ensure accurate monthly volume accounting, concurrency management, and audit integrity, even after the underlying tabular file content has been completely purged.

6. Security Safeguards & Encryption

Rowforma implements technical and organizational safeguards commensurate with the sensitivity of tabular data processing:

  • Encryption in Transit: All HTTP interactions with web portal, API endpoints, and embedded importer components require TLS encryption (HTTPS).
  • Cryptographic API Key Hashing: Server-side API key secrets (rf_live_...) are displayed to the user once upon creation. Rowforma stores only a one-way SHA-256 cryptographic hash of the key. Plaintext keys cannot be recovered by Rowforma staff or through database compromise.
  • AES-256-GCM Webhook Secret Encryption: Customer-configured webhook signing secrets are encrypted at rest in the PostgreSQL database using authenticated AES-256-GCM cryptography.
  • Multi-Tenant Tenant Isolation: Every database query and storage lookup is scoped to the authenticated user’s verified accountId context. Access tokens and session identities are validated per-request on the server.

7. Cookies and Browser Storage

Rowforma maintains a minimal browser storage footprint. We do not use advertising, marketing, or cross-site tracking cookies:

MechanismNameType / LifetimePurpose
HTTP Cookie__sessionStrict / Session (max 5 days)Essential authentication session token issued upon sign-in. Configured as HttpOnly, Secure, and SameSite=Lax.
Local Storagerowforma-themePersistent in browserRemembers your user interface color preference (light, dark, or system). Contains no personal data.

8. Subprocessors & Hosting Environments

Rowforma clearly distinguishes active operational infrastructure from optional or planned deployment targets:

A. Default Local Development & Preview Environment

In the local development environment, all database records reside in local PostgreSQL (127.0.0.1:5432), authentication is managed via a local Firebase Auth emulator (127.0.0.1:9099), and files are stored on local disk storage. No external transmission occurs.

B. Optional Serverless Database (Neon)

When explicitly configured via environment variable, persistent database records may be hosted on Neon Serverless Postgres (Neon, Inc.). Neon provides managed PostgreSQL hosting with infrastructure hosted in cloud provider regions (typically AWS US-East).

C. Authentication Provider (Google Firebase)

In production preview configurations, account authentication and identity tokens are managed through Google Cloud Firebase Authentication (Google LLC), subject to Google’s security standards and privacy disclosures.

D. Planned Infrastructure Notice (Not Operational)

Configurations for automated Google Cloud Run application hosting and Google Cloud Storage bucket persistence exist within the deployment codebase but are not currently deployed or operational. Rowforma does not list non-operational services as active processors.

9. Cross-Border Data Processing Notice

Notice under Alberta PIPA Section 13.1 and Canadian Privacy Standards:

Rowforma and its cloud service providers (such as Google Firebase or Neon) may store or process account information, system metadata, or data files on servers located outside of Canada, predominantly in the United States of America.

When information is processed outside Canadian borders, it becomes subject to the laws of the foreign jurisdiction in which it is held. Courts, law enforcement, and national security authorities in those jurisdictions may access such data pursuant to lawful legal process under local statutes. For questions regarding foreign service provider practices, contact our privacy office at the contact details provided below.

10. Embedded Importer Privacy Boundaries

When third-party software applications embed the Rowforma CSV importer via our JavaScript SDK or iframe:

  • End users submitting data through an embedded importer are not Rowforma account holders. Their legal relationship is with the host application.
  • Rowforma processes submitted CSV data strictly as a data processor for the host customer account that created the embed session.
  • Embed sessions validate the host application’s origin domain before exchanging messages (window.postMessage).
  • The host application remains solely responsible for providing appropriate privacy disclosures and obtaining necessary consent from its end users prior to ingestion.

11. Deletion, Access, and Correction Rights

Under applicable Canadian data privacy laws, you have rights regarding your personal information:

  • Access and Correction: You may request access to personal information we hold about you, or request correction of inaccurate profile data, by submitting a written request to our privacy office.
  • Account Deletion Limitation: The current Rowforma software release does not feature an automated self-service account deletion button in the settings portal. To request manual account termination and purge of associated profile identifiers, please submit a written request.
  • File Deletion: Uploaded datasets expire and are purged automatically within 24 hours without requiring a manual deletion request.

12. Operator Contact & Privacy Enquiries

If you have questions, concerns, or requests regarding this Privacy Policy or our information handling practices, please contact:

Operating Name: rowforma

Legal Entity Name: Not yet configured (Development Preview)

Privacy Office / Inquiries: contact@bazeflow.com

General Support: contact@bazeflow.com

If your privacy concern is not satisfactorily resolved, you have the right to contact the Office of the Information and Privacy Commissioner of Alberta (www.oipc.ab.ca) or the Office of the Privacy Commissioner of Canada (www.priv.gc.ca).