Coupa Software Incorporated

In an embodiment, a method includes receiving training data representing historic consumer demand for products, detecting changepoints in that data that may be associated with disruptive events, identifying relevant data for modeling, performing clustering, processing configuration information, training one or more machine learning models that are capable of evaluating other received data more accurately, and outputting results to a user display device.

In one embodiment, a method can access an expense line for a travel expense for an enterprise. The expense line is the subject of a procurement action under a relevant contract. The method can use a first machine learning model, determining a category associated with the expense line. The method can evaluate the expense line using the category associated with the expense line and a table of categories that are procurable with references to related travel contracts. The method can determine a match between a description field of the expense line and description fields of historic invoice records associated with travel contracts. In response to determining the match, the method can execute a responsive action for the expense line, wherein the responsive action includes generating a notification to a user, marking the expense line, and initiating a workflow to apply the contract to the expense line.

Input data comprises an order to provide a quantity of one or more finished goods to a site. Software is programmed for: accessing information that defines an architecture of a supply chain network comprising a plurality of sites connected by a plurality of path fragments; generating one or more path solutions to fulfill the order; removing one or more loops from the one or more path solutions; generating based on the one or more path solutions and the one or more removed loops, one or more revised path solutions; outputting the one or more revised path solutions for display on a client device; and transmitting, only to computers of the sites of a revised subset of a plurality of path fragments, a plurality of signals to initiate movement of one or more of raw materials or finished goods.

Embodiments of the disclosed technologies provide solutions for automatically reading digital electronic documents that contain tables and correctly extracting table data, rows and columns from the documents with high accuracy and high throughput. Embodiments are capable of converting a table portion of a read-only document to a searchable, editable data record using text rectangle (TR)-level numerical data that indicates probabilities of TRs belonging to canonicals and at least one convolutional neural network (CNN) that processes the TR-level numerical data to produce table-level numerical data.

In one embodiment, a method includes detecting activity of one or more agents that modifies a processing state of a first of multiple invoices, each invoice requesting approval. The activity is conducted by the one or more agents through interacting with the procurement processing system. The method includes updating, based on the detected activity, an invoice interaction history record associated with the first invoice in an invoice interaction history datastore communicatively coupled to the procurement processing system. The method includes generating one or more metrics regarding activity of the agents affecting a processing state of the invoices based on the updated invoice interaction history datastore. The method includes providing the one or more metrics for presentation to the procurement processing system.

Embodiments provides a method executed by a server computer executing a supply chain network analysis application of a supply chain model. The method includes receiving supply chain network data associated with a supply chain network having one or more supply chain nodes. The method then programmatically executes inferences on the supply chain network data using one or more machine learning models and one or more heuristic algorithms to implement descriptive analytics, diagnostic analytics, and prescriptive analytics to create and store one or more scenario prescriptions that specify one or more changes to the one or more supply chain nodes. The method performs steps for programmatically executing inferences on the supply chain network data that includes extracting one or more data features at a path level, the one or more data features indicating descriptive insights related to one or more paths in the supply chain network. The method includes a step of identifying, using one or more path-level machine learning models, one or more cost drivers of the one or more paths in the supply chain network by computing a feature score of each of the one or more data features at the path level. The method includes creating and storing, using the one or more path-level machine learning models and the feature score of each of the one or more data features, one or more digital representations of the one or more scenario prescriptions. The method includes generating and displaying one or more visualizations of one or more updated network models that implements the one or more scenario prescriptions.

In an embodiment, a computer-implemented method comprises storing, in one or more data repositories, transactional data relating to past transactions involving a plurality of commodities between a plurality of buyer entities and a plurality of supplier entities; calculating metric data of one or more overages from the transactional data, wherein each of the one or more overages indicates an amount to which a cost of a commodity item as specified in an invoice exceeds a cost of the commodity item as specified in a requisition or a purchase order, the invoice being generated in response to the requisition or the purchase order; receiving, from a computer associated with a buyer entity, a request to generate a requisition or a purchase order for one or more commodity items; identifying, from the request, a supplier identification (ID) associated with a particular supplier entity of the plurality of supplier entities or a commodity ID associated with a particular commodity of the plurality of commodities, for a particular commodity item of the one or more commodity items; calculating a projected total cost for the particular commodity item from the metric data using the supplier ID or the commodity ID, the projected total cost indicating a probable cost for the particular commodity item based on previous transactions that included the particular commodity item and particular supplier entity; in response to determining that the projected total cost for the particular commodity item is less than a threshold value, transmitting an approval of ordering the particular commodity item.

In an embodiment, a computer-implemented method includes parsing a communication record, determining one or more intents of the communication record using a trained machine learning model, linking the communication record to one or more record IDs, identifying one or more fields within the communication record, presenting the communication record in a standardized form, including identification of intent, the standardized form resulting from the identified one or more fields, and determining a corresponding intent related action on the determined one or more intents.

A computer-implemented method of detecting account codes and displaying the detected account codes on a graphical user interface comprising receiving, by a recommendation engine of a recommendation system, invoice data comprising supplier-customer information that corresponds to a supplier-customer transaction, wherein the invoice data comprises invoice descriptions and invoice characters, wherein the invoice descriptions and the invoice characters define contexts and patterns; determining, by the recommendation engine, that an amount of the invoice characters is not more than a preset threshold number of characters; in response to determining that the amount of the invoice characters is not more than the preset threshold number of characters, filtering, by the recommendation engine, the invoice descriptions of the invoice data based on predetermined constraints to extract filtered invoice data comprising filtered description lines and to generate a training corpus for a pre-trained Natural Language Processing (NLP) model; identifying, by classifying the filtered description lines with the pre-trained NLP model, one or more categories associated with each of the filtered description lines of the filtered invoice data; matching, by the recommendation engine, each of the identified one or more categories with one or more predefined historical categories, wherein the contexts and patterns associated with the filtered description lines are matched with predefined contexts and patterns of predefined historical invoice data that corresponds to the same supplier-customer information; generating, by the recommendation engine, a feature vector for the invoice data based on the matching; computing, by the recommendation engine, a categorical similarity score for each of the identified one or more categories based on the feature vector and an additional feature vector, wherein the additional feature vector is based on the predefined historical invoice data; and displaying, by the recommendation engine on the graphical user interface, a recommendation including an account code based on the computed categorical similarity score of each of the one or more categories to map the account code to the invoice data.

In an embodiment, a method includes receiving training data representing historic consumer demand for products, detecting changepoints in that data that may be associated with disruptive events, identifying relevant data for modeling, performing clustering, processing configuration information, training one or more machine learning models that are capable of evaluating other received data more accurately, and outputting results to a user display device.