System for dynamic effort-based delivery value predictive updates
Described are systems and processes for generating dynamic effort-based delivery value predictions for real-time delivery of perishable goods. In one aspect, a system is configured for generating dynamic delivery value predictions for delivery opportunities provided to couriers. For each order, delivery events and corresponding timestamps are received from devices operated by customers, restaurants, and couriers. Based on the timestamps, the system generates a predicted delivery duration with trained predictive models that use weighted factors such as order data and historical restaurant data. A service value for the delivery of the order is determined based the predicted delivery duration and a predetermined active time value. The service value is then transmitted along with the corresponding delivery opportunity to a user device of a courier. The determined service values may be adjusted based on courier acceptance rates of delivery opportunities and other factors such as customer experience.
1. A method comprising:
training a predictive model to output an active time value using a first dataset, the first dataset including a first set of pairing responses and parameters associated with orders corresponding to the first set of pairing responses;
receiving a second set of pairing responses associated with pairing notifications transmitted to a plurality of courier devices, wherein each pairing notification includes a service value based on the active time value and a respective predicted delivery duration, wherein the second set of pairing responses indicates a courier acceptance rate; and
inputting a second dataset into the predictive model to generate an updated active time value when the courier acceptance rate is above or below a predetermined threshold range, wherein the second dataset comprises the second set of pairing responses and parameters associated with orders corresponding to the second set of pairing responses.
2. The method of claim 1 , wherein the predictive model generates the updated active time value for a location associated with the orders corresponding to the second set of pairing responses.
3. The method of claim 1 , wherein the predictive model generates the updated active time value based on the courier acceptance rate associated with the second set of pairing responses.
4. The method of claim 1 , wherein the updated active time value is dynamically adjusted based on a time of day.
5. The method of claim 1 , further comprising generating a supplemental value amount based on a plurality of weighted factors, wherein the service value is adjusted by the supplemental value amount.
6. The method of claim 1 , wherein the predictive model employs a gradient boosted machine.
7. The method of claim 1 , wherein the predictive model is continuously trained using additional datasets comprising delivery parameters associated with subsequent orders.
8. The method of claim 1 , wherein the parameters associated with orders corresponding to the first set of pairing responses includes one or more of the following:
sub-totals of the respective orders corresponding to the first set of pairing responses,
types of items in the respective orders corresponding to the first set of pairing responses,
a number of items in the respective orders corresponding to the first set of pairing responses, and
a number of pending orders at associated restaurants.
9. The method of claim 1 , wherein the parameters associated with orders corresponding to the first set of pairing responses are received from one or more of the following: the plurality of courier devices, a restaurant device, and a customer device.
10. A system comprising:
one or more processors;
memory; and
one or more programs stored in the memory, the one or more programs comprising instructions for:
training a predictive model to output an active time value using a first dataset, the first dataset including a first set of pairing responses and parameters associated with orders corresponding to the first set of pairing responses;
receiving a second set of pairing responses associated with pairing notifications transmitted to a plurality of courier devices, wherein each pairing notification includes a service value based on the active time value and a respective predicted delivery duration, wherein the second set of pairing responses indicates a courier acceptance rate; and
inputting a second dataset into the predictive model to generate an updated active time value when the courier acceptance rate is above or below a predetermined threshold range, wherein the second dataset comprises the second set of pairing responses and parameters associated with orders corresponding to the second set of pairing notification responses.
11. The system of claim 10 , wherein the predictive model generates the updated active time value for a location associated with the orders corresponding to the second set of pairing responses.
12. The system of claim 10 , wherein the predictive model generates the updated active time value based on the courier acceptance rate associated with the second set of pairing responses.
13. The system of claim 10 , wherein the updated active time value is dynamically adjusted based on a time of day.
14. The system of claim 10 , wherein the one or more programs further comprise instructions for generating a supplemental value amount based on a plurality of weighted factors, wherein the service value is adjusted by the supplemental value amount.
15. The system of claim 10 , wherein the predictive model employs a gradient boosted machine.
16. The system of claim 10 , wherein the predictive model is continuously trained using additional datasets comprising delivery parameters associated with subsequent orders.
17. The system of claim 10 , wherein the parameters associated with orders corresponding to the first set of pairing responses includes one or more of the following:
sub-totals of the respective orders corresponding to the first set of pairing responses,
types of items in the respective orders corresponding to the first set of pairing responses,
a number of items in the respective orders corresponding to the first set of pairing responses, and
a number of pending orders at associated restaurants.
18. The system of claim 10 , wherein the parameters associated with orders corresponding to the first set of pairing responses are received from one or more of the following: the plurality of courier devices, a restaurant device, and a customer device.
19. A non-transitory computer readable medium storing one or more programs configured for execution by a computer, the one or more programs comprising instructions for:
training a predictive model to output an active time value using a first dataset, the first dataset including a first set of pairing responses and parameters associated with orders corresponding to the first set of pairing responses;
receiving a second set of pairing responses associated with pairing notifications transmitted to a plurality of courier devices, wherein each pairing notification includes a service value based on the active time value and a respective predicted delivery duration, wherein the second set of pairing responses indicates a courier acceptance rate; and
inputting a second dataset into the predictive model to generate an updated active time value when the courier acceptance rate is above or below a predetermined threshold range, wherein the second dataset comprises the second set of pairing responses and parameters associated with orders corresponding to the second set of pairing responses.
20. The non-transitory computer readable medium of claim 19 , wherein the parameters associated with orders corresponding to the first set of pairing responses includes one or more of the following:
sub-totals of the respective orders corresponding to the first set of pairing responses,
types of items in the respective orders corresponding to the first set of pairing responses,
a number of items in the respective order corresponding to the first set of pairing responses, and
quantities of pending orders at associated restaurants.