Service Robo is a asynchronous analyst who scans through the history of a customer and posts suggestions based on the current mouse click. Example in the below online grocery store application, repeated customer started purchasing grocery and forgot to add toothpaste, our ServiceRobo knows that the usage history of a particular customer and started posting "Hello Customer, looks like you missed adding toothpaste to your cart, do you want to consider it, and we have a offer too in this month, your accumulated amount after adding tooth paste is going to be $220.
Shopping assistant should be a guide through shopping based on customer purchase history, better not to average on 1 month data and start judging customer needs.
Current Offers
Shopping assistant also should be able to highlight the current offers and benefits once offer is utilized.
Regular Items
Shopping assistant will be ready with regular item list once customer clicks on menu bar and runs through last month receipt, it will help customer not go through menu bar and adding more frequent used grocery to the cart.
Properties of a selected Item
Once customer selects item, our agent will show small placard containing all the item properties like expiry date, calories and how many such items customer has consumed so far and so on.
Thought Share
1) Link Window shopper to E-commerce with face detection software and keep suggesting on windows shopped products.
2) Purchase history of a customer will give
3) Popup adds based on customer habits and body type (e.g. Customer is a fatty body personality, slimming programs and low calories food items.
4) Nuaral networks should be build to share customer information among the eCommerce websites, e.g. Customer has visited apparels website and bought cloths and later logged in to foot ware online, then our service agent should start guiding on matching to the apparel he/she bought, but needs to be again apply analytical theory and get his/her interesting color and materials, most importantly current trends if its already with our super computers. Customer should get a feel of his/her friend guiding through out shopping.
Customer information should be kept with Internet Service Providers and provide it as a secured transmission streams to interested parties with proper authentication.
Design consideration
Agent is a feasible add on to the application if we have a big data persistence and very good analytical mechanism in place as a wrapper service. Assistant will fire a query on each mouse click and sends feed to our analyst wrapper service, service should fetch history and apply artificial intelligence and return suggestions to agent. More to come in this space, I have just started blogging my thoughts.