Automatic order generation using trained model
A service provider system may store buyer information and merchant information related to past purchasing history of one or more buyers. In some examples, based in part on the buyer information and/or the merchant information the service provider system may identify items that the buyer regularly or habitually purchases. In some cases, the service provider system may pre-order or pre-purchase the items on behalf of the buyer and send a message to a device associated with the buyer to inform the buyer that an order for the regularly purchased item has been placed.
1 . A computer-implemented method for predictive ordering, the computer-implemented method comprising:
storing, at a data store, a buyer profile corresponding to a buyer, wherein the buyer profile includes one or more trained models corresponding to the buyer, wherein the one or more trained models corresponding to the buyer are trained using training data indicative of at least a purchasing history of the buyer;
predictively generating, by a computing device and using the one or more trained models corresponding to the buyer, a specified order for a specified time based on a trend identified in the purchasing history of the buyer, wherein the one or more trained models identify the trend based on an interval with which the buyer makes purchases according to the purchasing history;
identifying, by the computing device, a merchant to fulfill the specified order; and
automatically placing, by the computing device, the specified order with the merchant for the specified time on behalf of the buyer.
2 . The computer-implemented method of claim 1 , further comprising:
generating, by the computing device, a confidence rating indicating a likelihood of the buyer to place the specified order at the specified time; and
determining, by the computing device, that the confidence rating exceeds a predetermined confidence threshold, wherein predictively generating the specified order for the specified time is based on the confidence rating exceeding the predetermined confidence threshold.
3 . The computer-implemented method of claim 1 , further comprising:
training, by the computing device, the one or more trained models corresponding to the buyer using the training data.
4 . The computer-implemented method of claim 1 , further comprising:
sending, from the computing device to a buyer device corresponding to the buyer, a request for the buyer to confirm the specified order; and
receiving, at the computing device from the buyer device, a confirmation of the specified order in response to the request, wherein the computing device automatically placing the specified order is responsive to receiving the confirmation of the specified order.
5 . The computer-implemented method of claim 1 , wherein the specified time includes a specified date that is determined based on the purchasing history of the buyer.
6 . The computer-implemented method of claim 1 , wherein predictively generating the specified order for the specified time is also based on at least one of an area that the buyer is in, a weather condition in the area that the buyer is in, or an event in the area that the buyer is in.
7 . The computer-implemented method of claim 1 , wherein predictively generating the specified order for the specified time is also based on at least one of a regularity of corresponding prior orders by the buyer, a frequency of corresponding prior orders by the buyer, or a recency of corresponding prior orders by the buyer.
8 . The computer-implemented method of claim 1 , wherein predictively generating the specified order for the specified time is also based on at least one of a calendar associated with the buyer, a time of day, a time of year, or a day of a week.
9 . The computer-implemented method of claim 1 , wherein the one or more trained models corresponding to the buyer are trained using the training data to identify times when the buyer tends to make purchases with regularity.
10 . The computer-implemented method of claim 1 , wherein the one or more trained models corresponding to the buyer are trained using the training data to identify times when the buyer tends to stray from a regular purchases.
11 . The computer-implemented method of claim 1 , wherein automatically placing the specified order with the merchant on behalf of the buyer includes automatically purchasing an item from the merchant on behalf of the buyer.
12 . The computer-implemented method of claim 1 , wherein automatically placing the specified order with the merchant on behalf of the buyer includes automatically scheduling an appointment from the merchant on behalf of the buyer.
13 . The computer-implemented method of claim 1 , further comprising:
sending, from the computing device to a buyer device corresponding to the buyer, a notification that the specified order has been placed with the merchant on behalf of the buyer.
14 . The computer-implemented method of claim 1 , wherein automatically placing the specified order with the merchant on behalf of the buyer includes sending a request to the merchant associated with the specified order and receiving a confirmation from the merchant in response to the request.
15 . The computer-implemented method of claim 1 , wherein at least a subset of the training data corresponds to purchases made by a second buyer other than the buyer.
16 . The computer-implemented method of claim 1 , wherein the one or more trained models include one or more trained statistical models.
17 . A system for predictive ordering, the system comprising:
a memory storing instructions; and
one or more processors that are configured to execute the instructions, wherein execution of the instructions by the one or more processors causes the one or more processors to:
store, at a data store, a buyer profile corresponding to a buyer, wherein the buyer profile includes one or more trained models corresponding to the buyer, wherein the one or more trained models corresponding to the buyer are trained using training data indicative of at least a purchasing history of the buyer;
predictively generate, using the one or more trained models corresponding to the buyer, a specified order for a specified time based on a trend identified in the purchasing history of the buyer, wherein the one or more trained models identify the trend based on an interval with which the buyer makes purchases according to the purchasing history;
identify a merchant to fulfill the specified order; and
automatically place the specified order with the merchant for the specified time on behalf of the buyer.
18 . The system of claim 17 , wherein the execution of the instructions by the one or more processors causes the one or more processors to:
generate a confidence rating indicating a likelihood of the buyer to place the specified order at the specified time; and
determine that the confidence rating exceeds a predetermined confidence threshold, wherein predictively generating the specified order for the specified time is based on the confidence rating exceeding the predetermined confidence threshold.
19 . The system of claim 17 , wherein automatically placing the specified order with the merchant on behalf of the buyer includes automatically purchasing an item from the merchant on behalf of the buyer.
20 . The system of claim 17 , wherein automatically placing the specified order with the merchant on behalf of the buyer includes automatically scheduling an appointment from the merchant on behalf of the buyer.