Data Handling Practices

This page describes, in practical terms, how data moves through MarketiQ AI — what we collect, where it lives, who can access it, how long we keep it, and how it is deleted. It is the operational companion to our Privacy Policy, GDPR Compliance statement, and Security documentation.

MarketiQ AI processes the following categories of data:

• Account & workspace data: Your name, email, company, role, workspace settings, team membership, and plan/billing status.
• Connected-account data: OAuth access tokens and the profile, page, audience, analytics, and content data that platforms (LinkedIn, X/Twitter, Meta, Google/YouTube, Google Analytics, Search Console, HubSpot, Shopify) share under the scopes you approve.
• Customer content: Posts, drafts, campaigns, decks, media assets, brand guidelines, research reports, and calendars you create or upload.
• Leads & CRM data: Contacts and leads you import (CSV), explicitly capture from the current page via the Chrome extension, or sync through official connected-platform/CRM APIs — names, titles, companies, profile URLs, and engagement history.
• Usage & diagnostic data: Log data, feature usage, device/browser type, and IP addresses used for security, debugging, and product improvement.

Ingestion. Data enters MarketiQ AI in three ways: (1) you type or upload it, (2) you authorize an OAuth connection and we fetch it from the platform's official API, or (3) you explicitly trigger the Chrome extension to capture visible information from the current page.

Processing. Inside the platform, data is processed to deliver the features you use — research, strategy generation, content drafting, scheduling, publishing, outreach, analytics, and reporting. AI features send only the minimum context needed to generate the requested output.

Egress. Data leaves the platform only when you direct it to: publishing a post through an approved official connected-platform workflow, sending a capability-authorized message on a supported non-LinkedIn channel, exporting a CSV/PDF report, or sharing a white-label portal with your client. LinkedIn personal outreach and social-inbox replies remain drafts in MarketiQ; you perform the final interaction in LinkedIn. We never sell data or share it with advertisers.

MarketiQ AI uses large language models and generative AI services to power research, strategy, content, and analysis features. Our practices:

• Purpose-bound: Your content and context are sent to AI providers only to fulfil the specific request you make (e.g., "draft this post"), never for unrelated purposes.
• No training on your data: We do not use your private content, leads, or connected-account data to train our own or third-party foundation models. We use API tiers of AI providers that contractually do not train on submitted data.
• Minimal context: Prompts include only the data needed for the task; credentials and access tokens are never included in AI prompts.
• Configurable models: Admins can review and configure which AI models the workspace uses from Dashboard → Settings → Models.

Data is stored and processed using the following categories of infrastructure and sub-processors, each bound by data processing agreements:

• Cloud hosting & databases: Application servers, databases, and object storage for media assets — encrypted at rest.
• AI model providers: LLM and generative media APIs used for the features described above.
• Email delivery: Transactional email (signup, password reset, notifications, reports).
• Payments: Subscription billing — card details are handled entirely by the payment processor and never touch our servers.
• Analytics & monitoring: Aggregate product analytics and error monitoring to keep the service reliable.

A current named list of sub-processors is available on request at admin@trayarunyaventures.com.

• Within your workspace: Role-based permissions (owner, admin, manager, editor, viewer) govern who can see and modify data. Client portal users see only the approvals and reports shared with them.
• Tenant isolation: Every record is scoped to an organization and workspace; cross-tenant access is technically prevented at the application layer.
• Within Trayarunya Ventures: Production access is limited to a small set of authorized engineers for operations and support, protected by strong authentication, and logged. Support staff access your workspace data only with your permission, to resolve your issue.

We retain each category of data only as long as it serves you or as the law requires:

• Account data: Life of the account; removed from the live system after verified deletion, with encrypted backups aging out within 30 days.
• Connected-account tokens: Until you disconnect the platform or delete your account. Local access is disabled immediately; supported OAuth and broker grants are then revoked with durable retries before the encrypted vault row is erased.
• Customer content & leads: Until you delete them in-app or delete your account.
• Usage & diagnostic logs: Rolling window, typically 90 days or less.
• Billing records: As required by tax and accounting law (typically 7 years).
• Backups: Encrypted backups are rotated on a fixed schedule; deleted data ages out of backups within the rotation cycle.

You can exercise control over your data at any time:

• In-app deletion: Delete individual posts, leads, assets, campaigns, or entire clients from the dashboard — deletion is immediate in the live system.
• Disconnect platforms: Settings → Integrations immediately disables local access and copied credentials. We revoke upstream grants where the provider offers a revoke API; static API keys must also be rotated or revoked in that provider's dashboard.
• Export: Dashboard → Trust & Privacy creates a comprehensive, machine-readable JSON bundle with credential material redacted and tenant scope enforced by role.
• Account deletion: Dashboard → Trust & Privacy verifies connector revocation, Stripe cancellation and object-storage erasure before deleting the live tenant. Shared organisations are preserved. If external cleanup cannot be confirmed, deletion pauses safely for retry or support.

The MarketiQ AI Chrome extension and the LinkedIn Copilot desktop app follow strict data-minimisation rules:

• The extension reads visible current-page context only after you explicitly invoke capture, audit, or drafting. It does not scan unattended lists, follow result links, or browse in the background.
• Captured data syncs to your own MarketiQ AI workspace and nowhere else.
• The desktop app keeps your LinkedIn session in an isolated browser profile on your machine; your LinkedIn credentials are never transmitted to or stored on our servers.
• LinkedIn personal outreach and interaction replies are human-operated. MarketiQ can research, prioritize and draft, but you perform every final Connect, Send, Like, Comment, InMail and profile-save action in LinkedIn's native UI. Separately, a post you explicitly schedule or publish to a connected account may be delivered through an approved official LinkedIn API. Planning limits and pause controls remain in place.

For any question about how your data is handled — including sub-processor lists, DPAs, or security reviews:

Trayarunya Ventures (MarketiQ AI)
Email: admin@trayarunyaventures.com
General: info@trayarunyaventures.com
Phone: +1 (971) 512-1701 (US) / +91-8954333390 (India)