Payman

Payman connects AI systems (via API webhooks or direct integration) to automatically generate payment instructions when AI identifies completed tasks. The system receives task metadata from AI pipelines, validates task completion criteria, and triggers deterministic payment execution without manual intervention. This eliminates the coordination gap between task identification and worker compensation by creating a direct data flow from AI output to payment processing.

Payman abstracts payment execution across multiple underlying payment processors (ACH, international transfers, crypto wallets, etc.) by maintaining a provider-agnostic payment routing layer. When a payment is triggered, the system selects the optimal settlement method based on worker location, currency, and configured payment preferences, then executes the transaction through the appropriate provider's API. This decouples task-to-payment logic from payment infrastructure details.

Payman provides an API endpoint that accepts task completion events from AI systems, validates the event payload against a configurable schema, and enqueues the event for payment processing. The validation layer checks for required fields (task ID, worker ID, amount), verifies amounts against configured limits, and detects duplicate submissions. Events are persisted to an audit log before payment execution, creating an immutable record of the task-to-payment chain.

Payman maintains a worker registry that stores payment method preferences, location, currency, and compliance status for each worker. The system allows workers (or admins) to register payment methods (bank account, crypto wallet, PayPal) and configure settlement preferences. When a payment is triggered, the system retrieves the worker's profile and uses the configured payment method for settlement. Profile updates are versioned and audited.

Payman generates reconciliation reports that match task completion events to payment executions, identifying successful payments, failed transactions, and pending settlements. The system provides transaction-level detail (task ID, worker ID, amount, provider, status, timestamp) and aggregate reporting (total paid, success rate, average settlement time). Reports can be exported in CSV or JSON format and integrated with accounting systems via API.

Payman allows administrators to define payment rules that enforce constraints on task-to-payment conversions, such as minimum/maximum payment amounts, per-worker daily/weekly limits, and task-type-specific rates. When a task completion event is received, the validation layer checks the proposed payment against these rules before execution. Rules are versioned and can be updated without redeploying the AI system, enabling dynamic compensation adjustments.

Payman emits webhook events to notify external systems of payment status changes (payment_initiated, payment_completed, payment_failed, payment_reversed). The system signs webhooks with HMAC-SHA256 to ensure authenticity and provides a webhook retry mechanism with exponential backoff for failed deliveries. External systems can register webhook endpoints and filter by event type, allowing them to react to payment events in real-time (e.g., update task status, notify workers, trigger accounting entries).

Payman accepts batch submissions of multiple task completion events in a single API call, processes them in parallel, and returns a batch result summary with per-task status. The system optimizes batch processing by deduplicating events, grouping payments by worker and provider, and executing settlements in bulk where possible. Batch processing reduces API overhead and improves throughput for high-volume labeling programs.

When a payment fails (due to insufficient funds, invalid account, network error, etc.), Payman automatically retries the payment with exponential backoff (1 minute, 5 minutes, 30 minutes, 4 hours, 24 hours). The system logs failure reasons and allows administrators to manually retry failed payments or adjust payment methods. Failed payments are tracked separately and can be queried via the reporting API, enabling visibility into payment issues.

Payman supports payments in multiple currencies by converting amounts at settlement time using real-time or fixed FX rates. The system allows administrators to configure FX rate sources (live market rates, fixed rates, or provider-specific rates) and tracks conversion rates for each transaction. Workers can specify their preferred payout currency, and Payman handles conversion automatically during settlement.