Every order question answered
from the actual order, not a guess

The role today

Order status questions are the most repetitive and the most urgent kind of customer message, and they usually land on the same support inbox as everything else, competing with complaints and product questions for the same person’s attention. A customer asking where their order is at 11pm on a Friday waits until Monday for an answer a database could have given instantly, and by the time a human replies, half the message is spent apologizing for the wait instead of answering the question.

What the agent takes over

The agent answers order status, tracking and estimated delivery by pulling the real order and shipment data at the moment of the question, across email, WhatsApp, Telegram, Instagram DM and the site’s own order lookup. It handles address corrections and item swaps within your stated window, checking real stock before promising a replacement rather than confirming something the warehouse cannot actually fulfill.

It processes cancellations within policy automatically and routes anything outside policy, a request past the cutoff, a partial refund, a damaged-item claim, to a person with the order history attached. Repeat contacts about the same order are recognized, so a customer who already asked once does not start the conversation from zero on their second message. A daily digest goes to your team showing which orders triggered an escalation and why, which becomes a running list of where the policy itself might need adjusting.

For stores selling across multiple channels, a marketplace, a website, a retail partner, the agent reads whichever order system the purchase actually came from rather than assuming everything lives in one place, so a customer who bought through a marketplace gets an answer as accurate as one who bought direct. It also recognizes when an order status question is really an early complaint, three messages in two days about the same delayed shipment, and raises the priority accordingly instead of treating each message as a fresh, unrelated question. This pattern recognition becomes one of the more useful signals a support team gets: which SKUs or carriers generate the most repeat worried contact, and therefore where to fix the underlying process instead of just answering around it.

What stays with humans

Refund amounts outside standard policy, disputed charges, and any customer who is escalating in tone go to a person. Your team owns the cancellation window, the swap policy and the stock thresholds the agent checks against; the agent enforces them, it does not set them.

Guards

Every order lookup and every change is logged against the order record itself, so there is a full audit trail per order, not just per conversation. Stock and shipment data are read live, never from a cache that could be stale. The agent runs in shadow mode against a batch of recent real orders before going live, and a kill switch stops it from making any further changes instantly if something looks off. Carrier and stock integrations are monitored for outages separately from the agent’s own uptime, so a courier’s API going down is caught and flagged rather than silently producing stale tracking answers that look current.

Price and timeline

“Agency runs it” keeps the agent on our infrastructure with a monthly support plan covering tuning and monitoring. “Full control, handover-ready” delivers the agent on your own servers and accounts with documentation, source and credentials, so your team can run and change it without us; it costs more up front because the handover package, your-infra deployment and internal documentation are built in from day one.

Related

See the AI agents overview for how we build and guard these systems, and Development for the services that often pair with this one. Inside the agents catalogue: Delivery status agent, Returns and refunds agent, Subscription management agent. For a narrower, single-process version of this work, see Order status answers, WhatsApp catalogue and order flow in the automations catalogue. For how the same guard patterns held up on real systems: ProBay marketplace AI agent team.

Want this running for your team? Get in touch and tell us where the messages pile up.