What could your agent handle for you?

Every month, the owner places an order from his supplier in France. This used to take about an hour — copying data between systems, checking quantities, formatting documents.

What the agent does:

1. Preps the order data from the local inventory system
2. Copies and formats the order to fit the supplier's ERP requirements
3. Catches errors in the supplier's confirmation documents
4. The owner still decides quantities (judgement call) — the agent handles everything else

Outcome: Order prep went from ~1 hour to ~30 seconds.

Online orders are fulfilled from the physical stores — staff pick and pack from shelves. Shopify's order printing wasn't good enough, so they used to print 10–50 orders by hand every morning.

What the agent does:

1. Polls Shopify for new orders every 5 minutes
2. Skips virtual-only items (gift vouchers, etc.)
3. Generates two PDFs per order: a packing slip with product thumbnails + a full order page with customer info, payment method, and a big PICKUP/DELIVERY banner
4. Emails each PDF to the store's printer automatically

Outcome: Zero manual printing. Orders flow straight to the printer as they come in.

The franchise website needed product pages in Hebrew (right-to-left) and French (left-to-right), with images, links, and translations. A lot was messy because of the two different languages.

What the agent does:

1. Edits Shopify pages directly via Telegram commands
2. Created ~100 product pages with proper RTL/LTR formatting
3. Fixed broken links and 404 errors along the way
4. Handles both Hebrew and French content in one workflow

Outcome: Entire website rebuilt and cleaned up — a project that had been sitting in the backlog for months.

Tea is imported by sea, and the agent tracks the actual vessel carrying each shipment with daily location updates.

What the agent does:

1. Monitors the vessel's real-time position
2. Sends daily location updates via Telegram
3. Flags unusual activity — e.g. the ship slowing down near Crete because it was optimising fuel to avoid arriving early and paying port waiting penalties

Outcome: The owner always knows where his shipment is — and why it's moving the way it is.

After an order is placed with the supplier, there's a full chain: confirm, wait for documents, follow up if they're late. The owner used to track all of this in his head.

What the agent does:

1. Monitors the full order lifecycle after placement
2. Reminds the owner to request updates 5 days after confirmation
3. Flags when documentation is missing or late
4. Tracks shipment status end-to-end

Outcome: Nothing falls through the cracks. Next step: the agent handles follow-up emails itself, with the owner in CC.

Employees work across two stores with different bonus structures, overtime rules, and location assignments. Overtime kicks in after 8h 36min per day. Bonuses are split pro-rata by hours worked at each store.

What the agent does:

1. Exports timesheets from EasyTeam (Shopify app)
2. Applies overtime rules per employee
3. Figures out who worked at which store
4. Pulls sales data and calculates bonuses based on each store's performance
5. Splits bonuses pro-rata by each employee's hours at that store vs total store hours
6. Outputs three Excel files: Hours, Bonuses, and Accountant Table (Hebrew format)

Outcome: A complex, error-prone monthly process handled in minutes.

Problems with clock-ins used to pile up until month-end. Now they're caught the same day while the info is still fresh.

What the agent does every evening:

1. Checks EasyTeam timesheets for the day
2. Clicks into every active employee individually (the overview table doesn't show in-progress shifts)
3. Reports: who worked, their times, open shifts (forgot to clock out), clock-ins after 6 PM, shifts longer than 12 hours, and swapped times

Outcome: Timesheet errors caught same-day instead of snowballing until payroll.

After building 100 pages, the agent audited its own work to make sure nothing was broken.

What the agent does:

1. Spins up sub-agents to crawl the entire site
2. Finds broken links and 404 errors
3. Fixes them automatically
4. The main agent QA's the sub-agents' output

Outcome: Self-auditing, self-fixing website maintenance. Next step: running weekly on a schedule.

Importing food means government paperwork. Every shipment needs import invoices matched with health certificates, sent as PDFs to the right contacts.

What the agent does:

1. Matches import invoices with corresponding health certificates
2. Assembles the required document package as PDFs
3. Sends them to the correct government contacts

Outcome: Compliance paperwork handled without manual document assembly.