Frequently asked questions

Pick one high-volume workflow and launch it in a single channel. Prove that answers are accurate and escalations reach the right people, then expand to more channels and packs.

General-purpose copilots are tied to one vendor's model. Threada is a work platform: every request becomes a tracked WorkItem with policy controls, governed execution, and outcome analytics — across any LLM.

No. Threada works alongside your current tools. It captures and structures requests, then executes approved actions in the systems you already run.

AI handles understanding and structuring. Policy controls, review queues, and approvals handle the exceptions. People stay in the loop for anything high-risk.

Containment, time-to-resolution, escalation quality, approval cycle time, and action success — outcomes on the record, not interaction counts.

Run evaluation gates that test extraction, routing, grounding, and action behavior. Roll out gradually with canary traffic, shadow mode, and one-click rollback.

A WorkItem is the central unit of work in Threada — a typed, lifecycle-managed object that captures intake, decisioning, execution, and outcome for every request.

Traditional message-only automation tools manage requests one message at a time. Threada manages work: intake, policy, approvals, actions, and outcome measurement, with audit trails and governance controls.

Yes. Most teams begin with a single workflow scoped to one channel or workspace and expand incrementally as confidence builds.

Websites, uploaded documents, API connectors, and feed-based sources. Each is registered as a managed knowledge asset with its own freshness and retention policy.

You set sync cadence and stale thresholds per source. Threada alerts on freshness drift and failed syncs so evidence stays current.

Yes. Citation requirements can be applied per workflow and intent. Low-evidence responses are routed to review or escalation automatically.

Yes. Decision steps and approval requirements are configuration-driven.

Yes. Channel-specific policy overlays can override baseline workflow behavior.

Yes. Kill switches can disable execution per channel, workflow, or integration scope.

Yes. Telemetry dimensions include both channel and workflow identifiers for cross-sectional analysis.

Yes. Execution telemetry includes integration and action dimensions with success, retry, and failure states.

Yes. Policy version is available as a first-class reporting dimension for quality and outcome analysis.

A request completed without manual handoff because policy and evidence supported a safe automated path.

Any managed document or synced source record that Threada keeps fresh for retrieval and evidence.

Yes. Anytime. Changes take effect next billing cycle.

You'll get warnings as you approach. There's a 7-day grace period to upgrade.

Save 20% with annual billing.

No. It's indexed for retrieval only.

99.9% target with auto-scaling and failover.

No. Threada enforces policy within AI work automation — it integrates with your existing GRC and identity tools rather than replacing them.

Policy overlays are versioned and scoped by tenant, channel, pack, and workflow. Every execution records the policy version that was active.

Yes. Lifecycle, policy, and action events are exportable for external audits, SIEM integration, and regulatory reporting.

99.9% monthly availability with explicit RTO (4h) and RPO (24h) objectives and documented incident runbooks.

Stripe subscriptions + usage counters; grace period automation (D-3/D-1 notices) and controlled disablement post grace.

Yes. Export feeds (GCS and BigQuery) and on-demand downloads are available for compliance reviews.

No. Threada is designed to orchestrate around existing systems through integrations and governed actions.

Yes. Approval requirements can be applied by workflow, channel, risk class, and action type.

Use EvalSuite gates for extraction, routing, grounding, and action behavior before broad rollout.

Yes. Web can be your first channel while you keep the same workflow primitives for future channels.

Yes. Integration execution is asynchronous and supports retries with auditable error classes.

Use correlation IDs, WorkItem history, decision events, and execution logs to reconstruct end-to-end flow.