The situation
A Southern California collectibles retailer that sells in-store, through live auctions, and on eBay spent most of every week on repetitive work. The work covered the path from finding inventory to shipping a completed order.
Growth meant hiring. The owner did not want to hire people for tasks that followed repeatable rules, but the business still needed a person to make decisions that depended on value, condition, or the needs of a customer.
The useful dividing line was not whether software could touch the work. It was whether a step followed a rule or required judgment. That distinction set the scope for the system.
- Researching inventory and bidding against the owner's pricing rules.
- Photographing items, identifying them, and creating listings one at a time.
- Reconciling stock across the store, live auctions, and eBay.
- Handling order intake, shipping administration, and buyer messages.
What we built
We connected custom software and a few off-the-shelf tools into one system. Each automated step followed the owner's rules, and the shared stock record kept the three selling channels tied to the same inventory information.
The system joined work that had previously been handled item by item. It monitored bidding, supported image capture and identification, generated listings and prices, synchronized stock, and moved order details into shipping and buyer communication without re-keying the same information.
The build treated the owner's pricing rules as instructions, not as a replacement for the owner's judgment. That boundary made it possible to automate repeatable work without pretending that every collectible could be evaluated by software.
- Automated bidding and monitoring against the owner's pricing rules.
- Image capture and item identification that generates listings and prices.
- One stock record synchronized across in-store sales, live auctions, and eBay.
- Order intake, shipping labels, and buyer messages without re-keying.
What stayed manual
Final buy decisions on high-value pieces stayed with the owner. Condition judgments also stayed manual because they needed a person's assessment. Customer service stayed with a person when the situation called for a human response.
These were intentional limits. The system handled steps that could follow a defined rule and left the decisions that needed context with the people responsible for the business.
- Final buy decisions on high-value pieces.
- Condition judgments.
- Customer service that needs a person.
The result
Roughly 100 to 200 hours of manual work removed per month, the equivalent of one to two people the business did not need to hire.