System for automated origination of capital based on point-of-sale data informed by location

1. A system for automated origination of capital, the system comprising:

a backend server coupled via the internet to a plurality of fixed and mobile terminals in establishments corresponding to subscribers to a point-of-sale (POS) subscription service; and

each of the subscribers employ the plurality of fixed and mobile terminals in conjunction with the backend server to take payments for orders that comprise processing of credit card transactions using an integral payment processor as a card, chip, or tap reader;

the backend server, configured to: process credit card transactions for each of the subscribers;

automatically deduct initial holdback amounts from processed credit card transactions based upon predicted total revenues over a payback period;

automatically deduct updated holdback amounts associated with each of the subscribers based upon updated predicted total revenues over the payback period;

employ historical POS data to predict future POS data corresponding to each of the subscribers, and to employ the future POS data to generate the predicted total revenues corresponding to each of the subscribers over the payback period, wherein the backend server comprises:

a recurrent neural network that is trained on an older half of the historical POS data as input to the recurrent neural network to predict future POS data, and that is executed using a younger half of the historical POS data to train outputs of the recurrent neural network;

update the recurrent neural network using an older half of updated historical POS data as input to the updated recurrent neural network to predict updated future POS data; and

generate the updated predicted total revenues by executing the updated recurrent neural network using new POS data corresponding to each of the subscribers.

2. The system as recited in claim 1 , wherein the holdback amounts are updated to account for variations in actual revenue versus predicted revenue for each of the subscribers.

3. The system as recited in claim 2 , wherein the holdback amounts are deducted daily.

4. The system as recited in claim 3 , wherein the establishments comprise restaurants.

5. The system as recited in claim 1 , wherein the backend server further comprises: a logistic regression engine, configured to perform a binary logistic regression on subscriber data and metrics derived from the historical POS data and the locations corresponding to each of the subscribers to generate prices for capital product offers extended to each of the subscribers, which are paid back to the POS subscription service via automatic deduction of the initial and updated holdback amounts.

6. The system as recited in claim 1 , wherein the payback period comprises 270 days.

7. A system for automated origination of capital, the system comprising:

a backend server coupled via the internet to a plurality of fixed and mobile terminals in establishments corresponding to subscribers to a point-of-sale (POS) subscription service; and

each of the subscribers employ the plurality of fixed and mobile terminals in conjunction with the backend server to take payments for orders that comprise processing of credit card transactions using an integral payment processor as a card, chip, or tap reader;

the backend server, configured to:

process credit card transactions for each of the subscribers;

automatically deduct initial holdback amounts from processed credit card transactions based upon predicted total revenues over a payback period;

automatically deduct updated holdback amounts associated with each of the subscribers based upon updated predicted total revenues over the payback period, said backend server comprising:

a revenue forecaster, configured to:

employ historical POS data to predict future POS data corresponding to each of the subscribers, and to employ the future POS data to generate the predicted total revenues corresponding to each of the subscribers over the payback period, wherein the revenue forecaster comprises:

a recurrent neural network that is trained on an older half of the historical POS data as input to the recurrent neural network to predict future POS data, and that is executed using a younger half of the historical POS data to train outputs of the recurrent neural network; and

update the recurrent neural network using an older half of updated historical POS data as input to the updated recurrent neural network to predict updated future POS data; and

generate the updated predicted total revenues by executing the updated recurrent neural network using new POS data corresponding to the each of the subscribers.

8. The system as recited in claim 7 , wherein the holdback amounts are updated to account for variations in actual revenue versus predicted revenue for each of the subscribers.

9. The system as recited in claim 8 , wherein the holdback amounts are deducted daily.

10. The system as recited in claim 7 , wherein the backend server further comprises: a logistic regression engine, configured to perform a binary logistic regression on subscriber data and metrics derived from the historical POS data and the locations corresponding to each of the subscribers to generate prices for merchant cash advance offers extended to each of the subscribers, which are paid back to the POS subscription service via automatic deduction of the initial and updated holdback amounts.

11. The system as recited in claim 7 , the payback period comprises 270 days.

12. A method for automated origination of capital, the method comprising:

a backend server coupled via the internet to a plurality of fixed and mobile terminals in establishments corresponding to subscribers to a point-of-sale (POS) subscription service; and

each of the subscribers employ the plurality of fixed and mobile terminals in conjunction with the backend server to take payments for orders that comprise processing of credit card transactions using an integral payment processor as a card, chip, or tap reader;

via the backend server, processing credit card transactions for each of the subscribers;

automatically deducting initial holdback amounts from processed credit card transactions based upon predicted total revenues over a payback period;

automatically deducting updated holdback amounts associated with each of the subscribers based upon updated predicted total revenues over the payback period;

employing historical POS data to predict future POS data corresponding to each of the subscribers, and employing the future POS data to generate the predicted total revenues corresponding to each of the subscribers over the payback period, wherein the employing comprises:

training a recurrent neural network using an older half of the historical POS data as input to the recurrent neural network to predict future POS data; and

executing the trained neural network using a younger half of the historical POS data to train outputs of the recurrent neural network;

updating the recurrent neural network using an older half of updated historical POS data as input to the updated recurrent neural network to predict updated future POS data; and

generating the updated predicted total revenues by executing the updated recurrent neural network using new POS data corresponding to the each of the subscribers.

13. The method as recited in claim 12 , wherein the holdback amounts are updated to account for variations in actual revenue versus predicted revenue for each of the subscribers.

14. The method as recited in claim 13 , wherein the holdback amounts are deducted daily.

15. The method as recited in claim 14 , wherein the establishments comprise restaurants.

16. The method as recited in claim 12 , further comprising: performing a binary logistic regression on subscriber data and metrics derived from the historical POS data and the locations corresponding to each of the subscribers to generate prices for capital product offers extended to each of the subscribers, which are paid back to the POS subscription service via automatic deduction of the initial and updated holdback amounts.

17. The method as recited in claim 12 , wherein the payback period comprises 270 days.