E-commerce

E-commerce with AI agents: create products, analyse orders, reorder.

An online shop is made of many small jobs: taking over products, analysing orders, checking stock, reordering, matching invoices. An agent does this by your rules and asks you before anything goes out. Here is a Monday with four such runs.

A Monday in the online shop

Four agent runs in one day. Each connects several systems and handles a whole process, not just one step.

  1. 01 · 02:00

    Knowing your stock and reordering

    Reorder every two weeks without counting. Stock arrives from the warehouse at night, the agent calculates demand and sends the order after your yes.

    1. 1Stock arrives in records every night
    2. 2Every two weeks the agent calculates demand
    3. 3One yes from you, then the order goes out
    4. 4Order sent, open items created
    Warehouse systemRecordsSuggested orderSlackEmail to supplier
  2. 02 · 07:00

    New supplier products into the shop

    Your supplier releases new products. One agent run finds them, pulls data and images, creates them in Shopify and asks you before publishing.

    1. 1The supplier has new products
    2. 2All product data ends up in one record
    3. 3Supplier photos become product images
    4. 4The product is a draft in Shopify
    5. 5One yes in Slack, then it goes live
    ScheduleSupplier siteRecordImage AIShopifySlack
  3. 03 · 08:00

    Analysing orders and returns

    You want to know what comes back and why. Orders flow into records on their own, the report arrives every Monday, with a suggestion.

    1. 1Every order becomes a record
    2. 2Returns end up on the same order
    3. 3The returns report writes itself
    4. 4The agent suggests what to change
    Shopify webhookRecordsReportSlackShopify
  4. 04 · 09:10

    Goods received, invoice and credit note

    Does the invoice match what arrived? The agent compares order, invoice and goods received and tracks the credit note.

    1. 1The supplier invoice arrives by email
    2. 2What really arrived?
    3. 3You are informed, and so is the supplier
    4. 4Did the credit note arrive?
    InboxRecordWarehouse systemSlackEmail to supplierInbox

Common questions

Only if you allow it. The default: the agent creates drafts, you confirm, then they go live. For things you trust (for example price changes below 5%) you can switch the approval off.

Shopify, Klaviyo, Meta Ads, Gmail, Microsoft 365, IMAP inboxes, Slack, Google Drive and Google Sheets are ready to use. You connect a warehouse system, supplier portal or image AI via MCP or an HTTP step.

Yes. The agent reads the supplier's pages the way you see them in the browser and extracts products, prices and images. If the supplier changes the page, the run notices that details are missing and tells you instead of accepting gaps.

Yes, step by step: calculate demand, order, check goods received, match the invoice, track the credit note. Start with one step and add the next once the first one runs. At important points management gets a message or approves.

Describe the task. The rest takes shape on the platform.

Start for free, create your first skill with the assistant, and see what a run looks like before you connect anything.

Start for free