Privacy Policy & Data Protection Statement

2.2 Chat and Communications Monitoring

The Platform includes an integrated text-based messaging system enabling communication between Seekers, Providers, workers, and Platform administrators. All users are hereby expressly notified that they have no reasonable expectation of privacy in any communications transmitted through the Platform's messaging system. By using the Platform's messaging features, users explicitly consent to the monitoring, retention, and review of all messages as described in this section.

3.1 Three-Stage Automated Job Matching Pipeline

When a Seeker publishes a job on the Platform, an automated pipeline evaluates all registered Provider organizations to determine which Providers are eligible to receive a job offer. This pipeline operates in three sequential stages, with each stage capable of excluding a Provider from receiving an offer. No human reviews matching decisions — the process is fully automated. A Provider that is excluded by the matching pipeline will not receive a job offer and will have no opportunity to bid on the job unless the Seeker manually bypasses the pipeline.

3.2 Reliability Score Computation

The Platform maintains a Reliability Score for each Provider organization, expressed as a numeric value from 0 to 100. This score is computed automatically each time a relevant attendance event is recorded (including no-show detection) and is displayed to Seekers as a quality indicator. The Reliability Score may influence whether Seekers choose to extend job offers to a Provider and may be used as an input to future matching pipeline enhancements. The score is computed using the following weighted formula:

The Reliability Score is automatically recalculated each time a no-show is detected by the automated no-show detection cron job (see Section 3.4). Recalculation is triggered without any human intervention. Providers whose Reliability Score falls below defined thresholds may be deprioritized in matching results or, in extreme cases, reviewed by Crewd administrators for potential account suspension.

3.3 Dynamic Offer Time-to-Live (TTL) Assignment

When the matching pipeline generates a job offer for a Provider, the offer is assigned an automatic expiry time (Time-to-Live, or TTL) based on the proximity of the current date and time to the first shift date of the job. This TTL is assigned without human review. If the Provider does not accept or decline the offer before the TTL expires, the offer is automatically withdrawn and the Provider is excluded from consideration for that job. The TTL tiers are as follows:

• 48-Hour Offer Window: Assigned when the time remaining until the job's first shift is 120 hours (5 days) or more. Providers have 48 hours from the moment of offer delivery to respond.
• 24-Hour Offer Window: Assigned when the time remaining until the job's first shift is between 48 hours (2 days) and 120 hours (5 days), exclusive. Providers have 24 hours from the moment of offer delivery to respond.
• 12-Hour Offer Window: Assigned when the time remaining until the job's first shift is less than 48 hours (2 days). Providers have 12 hours from the moment of offer delivery to respond.
• These TTL thresholds are configurable by Crewd platform administrators via platform settings (stored under the key offerExpiryTiers). Any changes to these thresholds will be reflected in the Policy.

3.4 Automated No-Show Detection

The Platform operates an automated hourly cron job (a scheduled background process) that scans all confirmed and active jobs to detect potential worker no-shows. The cron job logic is as follows: for each scheduled shift in a confirmed or active job, if the current time is more than one (1) hour past the scheduled shift end time AND no attendance record exists AND no time entry exists for that shift, the system automatically creates an attendance record with the classification "no_show." This classification is applied without human review.

Upon automatic no-show detection, the following downstream automated actions are triggered without human involvement: (a) the Provider organization's Reliability Score is automatically recalculated (see Section 3.2); (b) automated notifications are dispatched to the Provider organization's administrators, the Seeker organization's representatives, and Crewd's internal operations team; and (c) a no-show penalty record is automatically created (see Section 3.6 regarding penalty types). Important: while the no-show record and penalty record are created automatically, no Stripe charge is executed automatically. All penalties require review and approval by a Crewd administrator before any financial charge is processed against a payment instrument.

3.5 Time Entry Auto-Confirmation

When a worker clocks out of a shift, a time entry record is created with a status of "pending" or equivalent, pending confirmation by the Seeker organization's representative. The Seeker organization has a window of twelve (12) hours from the clock-out timestamp to review and either confirm or dispute the time entry.

If the Seeker organization's representative does not take any action within the 12-hour review window, the time entry is automatically confirmed by the system without human action. Auto-confirmation has the same legal and financial effect as manual confirmation: the confirmed hours are treated as accurate and are used as the basis for billing calculations, including reconciliation invoices or credit notes.

This automated process may affect billing amounts. Seekers who believe a time entry is inaccurate should dispute the entry within the 12-hour window. After auto-confirmation, corrections require a formal dispute process with Crewd's support team.

3.6 Automated Penalty Creation (Four Penalty Types)

The Platform automatically creates penalty records in response to defined triggering events. There are four (4) penalty types. In all cases, the penalty record is created automatically, but no Stripe charge is executed without administrative review and approval.

• client_late (Seeker Late Cancellation Penalty): Automatically created when a Seeker cancels a confirmed job within twenty-four (24) hours of the job's first scheduled shift. The penalty amount reflects the agreed-upon liquidated damages for short-notice cancellation as specified in Crewd's Terms of Service.
• auto_nonpayment (Non-Payment Auto-Penalty): Automatically created when a job is cancelled due to non-payment — specifically, when a job has been in a paused state due to an outstanding invoice for seven (7) or more consecutive days and is subsequently auto-cancelled (see Section 3.8). The penalty is calculated based on the value of the cancelled engagement.
• trades_firm_withdrawal (Provider Withdrawal Penalty): Automatically created when a Provider organization withdraws from a confirmed job within the defined penalty window. The penalty window and amount are governed by the Terms of Service.
• trades_firm_no_show (Provider No-Show Penalty): Automatically created as a direct consequence of the no-show detection described in Section 3.4. Each automatically detected no-show generates a corresponding penalty record for administrative review.

3.7 Automated Organization Payment Blocking

The Platform automatically restricts a Seeker organization's ability to publish new jobs when the organization has one or more overdue invoices. Specifically, if a Seeker organization has an invoice in "overdue" status (meaning payment has not been received by the invoice due date), the system automatically sets a flag on the organization's account that prevents the organization from publishing new job postings.

The organization is automatically unblocked when all overdue invoices are marked as paid in the system. The transition from blocked to unblocked is automated and does not require manual intervention by a Crewd administrator, provided payment has been received and recorded by Stripe.

3.8 Automated Job Cancellation

The Platform operates automated cron jobs that cancel jobs in the following circumstances, without human review:

• Expired Published Jobs: A daily cron job (expiredJobsCron) scans all jobs in "published" status. If the job's first scheduled shift date has passed and the job has not been confirmed (i.e., no Provider has been matched and confirmed), the job is automatically cancelled.
• Paused Non-Payment Cancellation: A daily cron job (pausedJobsCron) scans all jobs in "paused" status. If a job has been paused due to non-payment for seven (7) or more consecutive days, the job is automatically cancelled. An auto_nonpayment penalty record is created upon such cancellation (see Section 3.6).
• In both cases, automated notifications are dispatched to all relevant parties (Seeker organization, Provider organization if one was matched, and Crewd operations) upon auto-cancellation.

3.9 Automated Billing Reconciliation

Following the completion of a job, the Platform's reconciliation system automatically compares the scheduled billing amounts (calculated at job confirmation based on scheduled hours and rates) against the actual confirmed hours (taken from confirmed or auto-confirmed time entries — see Section 3.5). This comparison produces one of three outcomes:

• Match: Scheduled and actual hours align within an acceptable variance threshold. No adjustment invoice is generated.
• Underpayment (Reconciliation Invoice): If actual confirmed hours exceed scheduled hours beyond the threshold, the system automatically generates a reconciliation invoice for the additional amount owed by the Seeker organization.
• Overpayment (Credit Note): If actual confirmed hours are less than scheduled hours beyond the threshold, the system automatically generates a credit note reflecting the overpayment amount owed to the Seeker organization.
• Reconciliation invoices and credit notes are created automatically by the system and are reflected in the Stripe invoice infrastructure. Financial impacts of reconciliation are determined algorithmically without human review unless a party formally disputes the outcome.