Generating an automatic packing list is no longer a luxury reserved for large companies with a six-figure ERP. Today, an SME that sells on marketplaces, manages dropshipping or distributes its own goods can build an intelligent workflow in a matter of days, not months. And the best part: no spreadsheets at three in the morning.
The real problem: every order channel is a source of errors
Orders that come in through different channels (email, WhatsApp, Shopify, WooCommerce, the PDF a sales rep sent…), a worker copying by hand, quantity errors, forgotten products and poorly utilised boxes. Multiply that by 50 shipments a day and you have a hidden cost that eats into your margin without showing up in any report. And the worst part isn't the time lost: what hurts most is the return of a product that never arrived or arrived wrong.
From chaotic Excel to the agent that prepares every box
A packing list is not just a list of products: it is the instruction that the person assembling the order in the warehouse follows. If that instruction arrives late, arrives wrong or simply doesn't arrive, the shipment fails. A system for generating automatic packing lists combines three pieces:
- A data source: the order, the delivery note, the customer's email, real-time stock.
- An AI agent that validates, transforms and decides what goes in each box.
- A clear output: printable PDF, label, message to the operator or directly to the warehouse team's chat.
When those three pieces are connected, the packing list stops being a static document and becomes a living process: it knows what is in stock, what is missing, what can be substituted and how to better group the day's orders.
4 steps to set it up even if you have no IT department
1. Centralise the orders into a single funnel. Decide where orders will come from to reach the agent: a dedicated email address (orders@yourcompany.com), a web form, a shared sheet or, most commonly, direct connectors to your online store. The less orders are touched before reaching the system, the fewer errors they will carry.
2. Document the rules of your current packing list. Before automating, write down what you currently do from memory. What columns does your list have? Do you group by customer, by route, by priority? What happens if a product is missing: do you substitute, wait or cancel? Those decisions, currently implicit in your operator's head, are what the AI agent has to learn. The more clearly you write them down, the better the system will perform.
3. Train the agent with your own history. A good agent is trained with real cases: 100–200 real packing lists, with their incidents already resolved. That way it learns that certain products always go together, that customer X prefers a specific box, or that on Mondays EU orders must be prioritised for customs reasons. You don't need thousands of examples; with 50–100 well-annotated ones the model already starts getting it right.
4. Connect the output to whoever executes it. The automatic packing list has to reach the right place: a printer in the warehouse, a tablet on the forklift, the logistics team's chat, or simply a QR label stuck to the box. That last mile is what turns automation into something you notice in your day-to-day work.
Concrete examples that work from day one
- Ecommerce with seasonal peaks. Black Friday, Christmas, sales season. An AI agent groups orders by postcode and generates a packing list per transport route, optimising loading time and reducing stops.
- B2B companies with recurring orders. If the same order from the big customer comes in every Monday, the agent detects it and suggests a pre-configured packing list, confirming with you any change in quantities.
- Dropshipping or marketplaces. The agent cross-references the order received with the supplier's stock and warns you before confirming the sale if there is a risk of running out. Your packing list reflects only what is actually available.
- Operations with substitutions. When an SKU runs out, the agent suggests alternatives based on your catalogue (same price, same category, same margin) and notes it on the packing list so the customer knows exactly what they will receive.
What you gain when you stop improvising
Fewer shipping errors and, therefore, fewer returns and refunds. Workers who stop keying in data and start preparing orders faster. Real traceability: every packing list is associated with its order, its customer and its shipment. And, above all, a clear picture of the bottleneck: is it picking, inventory, transport? At AizuaLabs we help logistics, ecommerce and distribution companies build these workflows with AI agents adapted to their real operations, not to a lab demo. If you want to stop generating packing lists by hand and start generating them properly, let's talk.
Frequently asked questions
What data does an AI agent need to generate automatic packing lists?
It needs access to your incoming orders, your catalogue or updated stock, and the rules you want to apply. With those three building blocks it can already produce a valid, traceable packing list.
Does it work with any ecommerce platform or ERP?
Yes, as long as there is a connector or channel (email, sheet, API) from which the agent can read the orders. It integrates with Shopify, WooCommerce, Holded, A3, Excel and many more.
How long does it take for the system to be operational?
It depends on the volume of history available, but a basic workflow is usually in production within 2–4 weeks, trained with your own packing lists and real cases.