Patent Grant
11562311
A portion of the disclosure of this patent document contains material which is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the Patent and Trademark Office patent file or records, but otherwise reserves all copyright rights whatsoever.
The present disclosure relates generally to robotic process automation, and more specifically to a system and method for identifying significant sequences of desktop events that have the highest return on investment (ROI) for robotic automation opportunities.
The subject matter discussed in the background section should not be assumed to be prior art merely as a result of its mention in the background section. Similarly, a problem mentioned in the background section or associated with the subject matter of the background section should not be assumed to have been previously recognized (or be conventional or well-known) in the prior art. The subject matter in the background section merely represents different approaches, which in and of themselves may also be inventions.
A computing device, such as a client device, may be used to enter input and perform various computing functions through mouse clicks and movements, keyboard and/or touch inputs, and other types of user input. As such, the user may repeatedly perform actions, which causes increased input and data processing by a computing device and leads to negative user experiences with data processing and software. Often various types of business processes that may be performed with a client device may require multiple different types of inputs and may be repeated with different input data, such as names, businesses, and other information. Businesses and other entities may wish to provide robotic automation of these processes and sequences of inputs so that users are not required to repeatedly perform such inputs. Robotic automation of these repeated processes provides savings in time, inputs, and processing resources. However, discovery and analysis of these robotic automation opportunities is presently done manually, and determination of ROI for automation opportunities may be difficult to determine and justify. This causes significant business process flows that can be automated to be missed. Additionally, this introduces bias and is time consuming and costly.
In the figures, elements having the same designations have the same or similar functions.
This description and the accompanying drawings that illustrate aspects, embodiments, implementations, or applications should not be taken as limiting—the claims define the protected invention. Various mechanical, compositional, structural, electrical, and operational changes may be made without departing from the spirit and scope of this description and the claims. In some instances, well-known circuits, structures, or techniques have not been shown or described in detail as these are known to one skilled in the art. Like numbers in two or more figures represent the same or similar elements.
In this description, specific details are set forth describing some embodiments consistent with the present disclosure. Numerous specific details are set forth in order to provide a thorough understanding of the embodiments. It will be apparent, however, to one skilled in the art that some embodiments may be practiced without some or all of these specific details. The specific embodiments disclosed herein are meant to be illustrative but not limiting. One skilled in the art may realize other elements that, although not specifically described here, are within the scope and the spirit of this disclosure. In addition, to avoid unnecessary repetition, one or more features shown and described in association with one embodiment may be incorporated into other embodiments unless specifically described otherwise or if the one or more features would make an embodiment non-functional.
Overview
For a desktop events conducted by a user, methods, data structures, and systems are provided for identifying and automating significant sequences of these desktop events based on ROI potential for automation. A database system stores tables and data structures of user desktop events that is accessible by an automation finder engine. The events may be generated, for example, by users or administrators (e.g., agents of an organization) of client systems or devices based on input and processing requests to the client devices, such as input and data while performing operations on the client devices.
The automation finder engine may process the desktop events and other user actions to split the user actions into sentences, or a string of user actions that acts as an entire input to perform some business process. These sentences of user actions act as a combination of several actions that express a particular business functionality. Using these sentences, repetitive sequences may be identified, which may be those sequences that have corresponding user actions that are consecutive and/or within the same time-frame and are repeated within a stream of user actions. The sequences may be filtered to identify the best ones of those sequences that have the highest ROI. Once significant ones are identified and named, those may be used to build one or more templates that permit entry of dynamic text when form filling or otherwise executing a business process.
The embodiments described herein provide methods, computer program products, and computer database systems for robotic process automation for significant sequences of desktop events using machine learning techniques. An automation finder system provides users with access to templates and other robotic automation processes based on identification of significant sequences for automation. For example, the online system may provide a portal that includes templates and other robotic automation processes. As part of providing the services to users, the online system stores these robotic automation processes and provides the processes for access through an automation finder portal. The robotic automation processes may be accessible, executable, and/or downloadable using the portal.
According to some embodiments, in a automation finder system accessible by a plurality of separate and distinct organizations, a machine learning model is provided for identifying, predicting, and automating sequences of user actions and desktop events based on ROI potential for automating such sequences, taking into account the specificities of each sequence and the sequences together, thereby enhancing the experience of users during use of client devices, reducing repeated input, and increasing processing speed and availability.
Example Environment
The system and methods of the present disclosure can include, incorporate, or operate in conjunction with or in the environment of an automation finder engine, database, and portal, which may include cloud-based computing architecture.
Clients 110 may be utilized in order to collect user desktop events and provide event logs and processing history associated with the desktop events. A desktop event may correspond to any desktop activity of a user, for example, a text copy through mouse clicks and/or keyboard commands, a text paste, a left or right mouse click, a selection of a tab command, input of text data, or other input and activity of the user. The desktop event may also be associated with particular user actions, which is the execution of the particular desktop event (e.g., the input) within a certain application, at a certain time, and/or according to another parameter (e.g., interface location, elements, etc.). These user actions may be joined into discreet sequences of actions, which define a particular execution of a process by a user. For example, a concatenation of several actions, which may or may not be consecutive but may be joined by a temporal similarity, may correspond to a complete business flow executed by a user, such as a process. These processes may therefore be identifiable and utilized for automation based on significant sequences of actions, or processes, that have a highest or best ROI for robotic automation.
The users of clients 110 may differ in their respective capacities, and the capacity of a particular client 110 might be entirely determined by applications, permissions, and profiles for the current user. For example, where a salesperson is using a particular client 110 to perform desktop events, that user system has the capacities allotted to that salesperson and those desktop events may be associated with a salesperson. However, while an administrator is using that client 110 to perform desktop events, that client 110 has the capacities allotted to that administrator and may perform desktop events associated with an administrator. Thus, different users will have different capabilities and perform different desktop events with regard to automation of desktop events, and corresponding profiles may be provided to server 120 when collecting and automating desktop events.
Each of clients 110 also typically include one or more user interface devices, such as a keyboard, mouse, trackball, touch pad, touch screen, pen or the like, for interacting with a graphical user interface (GUI) provided by the browser on a display (e.g., a monitor screen, liquid crystal display (LCD) monitor, light emitting diode (LED) monitor, organic light emitting diode (OLED) monitor, etc.) in conjunction with pages, forms, applications, and other information provided by the server 120 or other systems or servers. For example, the user interface device can be used to access data and applications hosted by server 120, and to perform searches on stored data, and otherwise allow a user to interact with various GUI pages that may be presented to a user. As discussed above, embodiments are suitable for use with the Internet, which refers to a specific global internetwork of networks. However, it should be understood that other networks can be used instead of the Internet, such as an intranet, an extranet, a virtual private network (VPN), a non-TCP/IP based network, any LAN or WAN or the like.
Clients 110 might communicate with server 120 and/or automation finder portal 140 using TCP/IP and, at a higher network level, use other common Internet protocols to communicate, such as hypertext transfer protocol (HTTP or HTTPS for secure versions of HTTP), file transfer protocol (FTP), wireless application protocol (WAP), etc. In an example where HTTP/HTTPS is used, client 110 might include an HTTP/HTTPS client commonly referred to as a “browser” for sending and receiving HTTP//HTTPS messages to and from an HTTP//HTTPS server, such as server 120. Such an HTTP/HTTPS server might be implemented as the sole network interface between server 120 and network 150, but other techniques might be used as well or instead. In some implementations, the interface between server 120 and network 150 includes load sharing functionality. Clients 110 may collect and provide user desktop events and actions and export them to server 120 as JavaScript Object Notation (JSON) or other file format over a secure connection, such as HTTPS.
Network 150 is any network or combination of networks of devices that communicate with one another. For example, network 150 can be any one or any combination of a local area network (LAN), wide area network (WAN), telephone network, wireless network, point-to-point network, star network, token ring network, hub network, or other appropriate configuration. As the most common type of computer network in current use is a transfer control protocol and Internet protocol (TCP/IP) network, such as the global inter network of networks often referred to as the Internet. However, it should be understood that the networks that the present embodiments might use are not so limited, although TCP/IP is a frequently implemented protocol.
According to one embodiment, server 120 is configured to provide webpages, forms, applications, data, and media content to clients 110 and/or to receive data from clients 110, including logs of desktop events used for determination of robotic automation opportunities of business processes. As such, server 120 provides security mechanisms to keep data secure. Additionally, the term “server” is meant to include a computer system, including processing hardware and process space(s), and an associated storage system and database application (e.g., object-oriented data base management system (OODBMS) or rational database management system (RDBMS)). It should also be understood that “server system” and “server” are often used interchangeably herein. Similarly, the database objects described herein can be implemented as single databases, a distributed database, a collection of distributed databases, a database with redundant online or offline backups or other redundancies, etc., and might include a distributed database or storage network and associated processing intelligence.
In some embodiments, the server 120, shown in
Several elements in the system shown in
Machine Learning Model
In order to provide automation opportunities of business processes, a machine learning model may be generated based on initial input data.
Memory 220 may be used to store software executed by computing device 200 and/or one or more data structures used during operation of computing device 200. Memory 220 may include one or more types of machine readable media. Some common forms of machine readable media may include floppy disk, flexible disk, hard disk, magnetic tape, any other magnetic medium, CD-ROM, any other optical medium, punch cards, paper tape, any other physical medium with patterns of holes, RAM, PROM, EPROM, FLASH-EPROM, any other memory chip or cartridge, and/or any other medium from which a processor or computer is adapted to read.
Processor 210 and/or memory 220 may be arranged in any suitable physical arrangement. In some embodiments, processor 210 and/or memory 220 may be implemented on a same board, in a same package (e.g., system-in-package), on a same chip (e.g., system-on-chip), and/or the like. In some embodiments, processor 210 and/or memory 220 may include distributed, virtualized, and/or containerized computing resources. Consistent with such embodiments, processor 210 and/or memory 220 may be located in one or more data centers and/or cloud computing facilities. In some examples, memory 220 may include non-transitory, tangible, machine readable media that includes executable code that when run by one or more processors (e.g., processor 210) may cause the one or more processors to perform any of the methods described further herein.
As shown, memory 220 includes automation finder engine 130. Automation finder engine 130 is implemented as a machine learning engine and/or neural network used to automatically analyze real-world information with human-like accuracy. In general, machine learning engines receive input information and make inferences and/or predictions using the input over time through a machine learning process, such as a supervised, unsupervised, and/or reinforcement learning model, although other types of learning may be utilized. Thus, a machine learning engine may proceed iteratively over an input corpus of documents or event file logs, such as win-32 logs of desktop events, and make similar inferences from the training examples that a human might make. Automation finder engine 130 may be used to implement and/or emulate any of the machine learning processes described further herein. According to some embodiments, automation finder engine 130 may be trained with one or more encodings or features so that it is able to predict and/or cluster similar desktop events into business processes for automation. This is described in more detail below.
After training of automation finder engine 130, in some embodiments, the machine learning model implemented by automation finder engine 130 is provided to automate desktop events for users, organizations, and companies executing such events, such as clients 110. Computing device 200 may receive or intake a user event log 250 of desktop events by a user, such as win-32 logs. User event log 250 can be unstructured data in that it can take the form of input fields, text, and actions. In some embodiments, the input sequence is not constrained, restricted, or required to be in a particular form or limited to a particular set of words, phrases, or vocabulary. An example of input according to some embodiments is shown in
The computing device 200, using the model of automation finder engine 130, generates related business process automation 260 for which an automation of a business process is performed according to a template, taking into account the significance and ROI of the business process. Business processes may be detected based on sequences of desktop events and user actions. An example of determination of a sequence of user actions is shown in
Low-level events are converted to natural text, for example, by removing unnatural text, such as computer programming and code, and/or converting the data to a format processable by automation finder engine 130. A sample data format of a desktop event and correspond user action is shown in
Further, each user action, such as user action 500a, is further associated with a “description” field that summarizes the action's information. This field acts as a unique identifier of the user action's functionality. This allows different actions to be associated with different descriptions while the same actions can be linked by the same or similar identifiers (e.g., the same or similar descriptions and identifiers within a vector space or other representation). In an exemplary embodiment, a description for user action 500a may be as follows: “description”: “User Click(LeftClick) on All applicationsFavoritesYour history(tab) in ABCD12345671 ServicePlease—Internet Explorer—iexplore”. Other descriptions and/or identifiers may be used instead.
During step 220 of preprocessing event logs, an additional preliminary step may split an input stream of user actions into “sentences” of user actions, or other user actions that are temporally linked so that the sentence describes a particular business functionality and/or process. For example, each sentence is taken from an input stream based on a time frame or period in which the user actions of the sentence occur. An exemplary conversion of a string of user actions split into sentences based on temporal similarity is shown in
In order to determine user action sentences 522, a time difference between subsequent user actions may be utilized, such that a high time difference between subsequent user actions implies that a user transitioned to another business functionality (e.g., execute a new different business process or repeated the previous business process). Thus, this may denote ending and beginning of a separate process. The time difference may be set by the system or an administrator, or learned based on time differences (e.g., analyzing differences between action time 506 and another time) in the specific user event log input and/or across all event logs. Additionally, the time difference may also be used with a specific time period or frame in which all input and desktop events are required to be detected so that sentences that are too long may be omitted. During the preprocessing step, the actions from the database may be processed and sentences may be identified by iterating the actions over the database of event logs. Actions may be identified separately so that each action may be identified with other actions occurring in temporal frequency and similarity to generate the particular user action sentences.
In order to train the machine learning model, grammar fixing and entity extraction may be performed to identify events embeddings that may be filtered for identification of business processes. At step 304, generalization of event descriptors may be performed such that noise within the description field (e.g., the field having the identifier that describes the user action's functionality) is reduced. For example, the description field may be noisy due to particular data that varies between user actions based on the input. Thus, even though the user action corresponds to the same functionality, a description may be different based on the local data input. In this regard, the local data input may be identified as variable input tokens, such as tokens that may include names, addresses, IP numbers, and other information that is particular to the certain input during a user action (e.g., a name added in a copy-paste user action). However, that user action (e.g., the copy-paste) may correspond to the same functionality and only have different dynamic input data.
For example, the following two descriptions of communications represent the same functionality of inserting text in a communicators conversation:
“User InputText on ISRAEL ISRAELI|conversation in communicator”
“User InputText on ISRAEL RANANA|conversation in communicator”
The descriptions above appear different, however, as they include different tokens for a NAME (e.g., “ISRAEL ISRAELI” and “ISRAEL RANANA”). In order to determine that these user actions correspond to the same user action, the unique token used individually for a “NAME” is replaced with more generalized descriptors of “NAME” so that the descriptions do not lose their identity and correspond to the same user action. In the above example, both would be replaced with: “User InputText on NAME I conversation in communicator”. This improves further processing efficiency to identify most important or significant business processes based on ROI by ensuring that user actions having the same functionality are represented in an identical manner even if their particular token data is different (e.g., due to different input of dynamic data for the action even when the business functionality of the action is the same). Other tokens in the description field may also be replaced depending on the particular dynamic data for the user action. The actions may be generalized by iterating a generalization process over the database of user actions and removing particular names and other tokens.
Once generalization of event descriptors and other tokens are replaced, event sequences may be mined, at step 306. Mining of the sentences occurs to find repetitive sequences, such as two or more user actions performed in sequence, which reoccur within the user action sentences. In sequential pattern mining, repetitive sequence in a data set that contains sets of sequences may be identified.
In order to perform mining of sequences from the user action sentences, the PrefixSpan algorithm may be utilized for sequential data mining. The PrefixSpan algorithm determines sequential patterns in data, such as sequential user actions in user action sentences, through a pattern growth method. The PrefixSpan algorithm may identify repetitive sequences having a minimum sequence support, and that minimum support may be tuned to determine particular sequences based on the data set itself. Additionally, a maximum support value may also be set and/or tuned using the data. When using the PrefixSpan algorithm, a sequence database may be converted into a smaller set of data and sequential patterns may be identified by analyzing locally frequent fragments within the data sets. Thus, the PrefixSpan algorithm may be used to find candidate sequences first and then recursively passing those candidate sequences through the data of user actions and desktop events to identify all sequences within the database.
The PrefixSpan algorithm is discussed in more detail in Mining sequential patterns by pattern-growth: the prefixspan approach, Pei J, Han J, Mortazavi-Asl B, IEEE Transactions On Knowledge and Date Engineering, 2004, 16(11):1424-1440, Analysis of Web Access Sequence Based on the Improved PrefixSpan Algorithm, Yang Xul and Yu Wang, International Industrial Informatics and Computer Engineering Conference (IIICEC 2015), and Mining Sequential Patterns Using I-PrefixSpan, Dhany Saputra, Dayang R. A. Rambli, Oi Mean Foong, World Academy of Science, Engineering and Technology 35 2007. However, other types of sequential data mining patterns may also be utilized including the GSP algorithm, SPADE algorithm, FreeSpan, or another algorithm.
An exemplary flowchart for sequence mining from data sets of user actions sentences is shown in
Using the PrefixSpan or other sequential pattern mining algorithm on the sentences, at step 406, the sequences are identified and returned in a sequences database. At step 408, it is determined whether the sequence database meets the minimum sequence length or size. For example, in
At step 408, if it is determined that the sequence database does not meet the minimum amount of number of sequences, then the support threshold may be tuned at step 410. Tuning the support threshold may include raising or lowering the support threshold based on whether lowering the support threshold causes the length of the sequence database to meet a particular sequence database size required by the system. For example, by lowering the support threshold, additional sequences may be identified but may cause the sequence database to exceed a set length or number of sequences. Increasing the threshold may further restrict sequences from the database but may cause the database to fail to meet a minimum length or number of sequences. This value may be tuned until the value falls within a range or amount required by the system. However, if the length of the sequence database does meet the minimum required size (e.g., has sufficiently identified sequences for business process automation), then the sequence database may be returned at step 412.
Referring back to
The scoring algorithm is shown at step 424 that scores each sequence, where the scoring algorithm is iterated over the sequences to determine a score for each sequence. The score for each sequence is calculated as score=norm_quality_score·ROI_score. In this regard, two rankings are weighted together to determine the score, which includes an ROI score and a quality score of the sequence. The ROI score corresponds to an amount of time that was spent on the sequence, such as a time from a starting timestamp of the first user action to a timestamp of the end user action and/or last data entry in the sequence. Thus, the ROI score identifies how much time was spent on the sequence and how much time can be saved by automating the sequence. The quality score represents a measure of the quality of the sequences actions. The quality of the sequences actions may associate a unique factor or value to each user action in the sequence based on the importance of the user action. For example, text input and/or copy-paste are more meaningful than a mouse-click or other action with no text. In some embodiments, this score may be normalized to a value between 1 and 1.5 (or any other value) so that the ROI score may be affected by a factor of up to 50% (or other amount, including devaluing the ROI score) based on the sequence's quality of actions.
At step 426, sequence filtering further includes sub-sequence removal to remove those sequences contained within a longer sequence and having a lower score than the longer sequence. For example, sub-sequence removal may proceed by filtering the sequences that are contained within a different sequence so that the higher scored sequences will be preferred. For example, consider a sequence having a-b-a-b-a-b of user actions in a sentence. Within that sequence, a-b (and others, such as a-b-a and b-a-b) may be found. However, a-b-a-b-a-b may have a higher score than a sequence of a-b (e.g., 10 vs. 5 after their scores are calculated using the aforementioned algorithm). Thus, the longer sequence may be preferred and be considered more significant for automation. Those sub-sequences are identified using the aforementioned process, at step 428. If such sequences exist, those sub-sequences are removed based on their score and the score of any parent sequences. Thus, when considering whether a sequence is significant for automation, or more significant than another sequence, the processes described herein may consider the sequence's ROI based on time for input and execution of the sequence by a user with additional factors. This scoring of sequences allows identification of a sequence's importance and significance, and further allows a quantitative measurement to be determined that allows comparison of sequences within a log of desktop events.
Clustering of sequences may further be performed to identify the same or similar sequences and select a sequence of those for automation. In this regard, sequence descriptions are converted to vector representation of the user actions, at step 430. Actions may be represented as vectors using word2vec or other technique to generate word vectors from user actions. For example, a vectorization process may generate a vector from word and/or text input of the user action. Each description and/or parts of the description can be a vector. In some embodiments, word vectors for actions may be attained by running methods like word2vec, FastText, or GloVe, each of which defines a way of learning word vectors with useful properties. In some embodiments, pre-trained vectors of a certain dimensionality may be used. In some embodiments, the embedding may include partial vectors related to portions of a word within the actions. For example, the word “where” includes portions “wh,” “whe,” “her,” “ere,” and “re.” Partial word embeddings can help to enrich word vectors with sub-word information. A vector representation of each sequence's description may then be generated through averaging the vectors of the descriptions included in the sequence.
At step 432, distance calculations are performed on sequences such that a distance between two sequences can be determined by weighing the Euclidean distance between vector representations of the two sequences with the Jaccard similarity between the two sequences. For example, a distance between sequence S1 and sequence S2 may be calculated as:
At step 434, using the distance between the sequences, clustering is performed to identify sequences with other similar sequences. A cluster represent sequences having similar business functionalities, such as sequences that implement the same or similar business processes for automation. When performing clustering, the following algorithm may be used to determine a modularity Q of a graph of those clusters within vector space:
This may be defined as a value between −1 and 1 and represent the density of the links within the individual clusters versus the links between the clusters. In the aforementioned formula, ki represents the sum of the distances for the i-th sequence to the rest of the sequences, ci is the cluster associated with the i-th sequence, and M is the sum of ki (i.e., M=Σki). Using this method, at step 436, an iterative finding of smaller clusters may be performed, wherein in each iteration, each sequence is associated with its cluster. Then for the sequence, the change in modularity is determined from its cluster. After determining the modularity for a sequence with its cluster, the sequence is then removed from the cluster and is associated with the cluster for its neighbor sequences. Iterating this over a database, the i-th sequence is associated with a cluster that maximizes the modularity.
Once clustering is performed, at step 438, the most significant sequences are returned. In each cluster, the sequence with the highest score is considered the most significant sequence and is returned for automation. The most significant sequences may then be named according to its actions. These significant sequences may then be analyzed for automation through scripts and templates allowing for dynamic data to be entered through one or more robotically automated processes.
For example, returning to
As discussed above and further emphasized here,
Some examples of computing devices, such as computing device 200, may include non-transitory, tangible, machine readable media that include executable code that when run by one or more processors (e.g., processor 210) may cause the one or more processors to perform the processes of method 300. Some common forms of machine readable media that may include the processes of method 300 are, for example, floppy disk, flexible disk, hard disk, magnetic tape, any other magnetic medium, CD-ROM, any other optical medium, punch cards, paper tape, any other physical medium with patterns of holes, RAM, PROM, EPROM, FLASH-EPROM, any other memory chip or cartridge, and/or any other medium from which a processor or computer is adapted to read.
Although illustrative embodiments have been shown and described, a wide range of modifications, changes and substitutions are contemplated in the foregoing disclosure and in some instances, some features of the embodiments may be employed without a corresponding use of other features. One of ordinary skill in the art would recognize many variations, alternatives, and modifications of the foregoing disclosure. Thus, the scope of the present application should be limited only by the following claims, and it is appropriate that the claims be construed broadly and in a manner consistent with the scope of the embodiments disclosed herein.
| Number | Name | Date | Kind |
|---|---|---|---|
| 20160219086 | Shukla | Jul 2016 | A9 |
| 20160350673 | Kumar | Dec 2016 | A1 |
| 20180113781 | Kim | Apr 2018 | A1 |
| 20200068031 | Kursun | Feb 2020 | A1 |
| Entry |
|---|
| R. Agrawal, et al.; Mining Sequential Patterns; ICDE '95 Proceedings of the Eleventh International Conference on Data Engineering; Mar. 6-10, 1995; pp. 3-14 (https://dl.acm.org/citation.cfm?id=655281 &preflayout=flat). |
| J. Pei, et al.; Mining Sequential Patterns by Pattern-Growth: The Prefixspan Approach; IEEE Transactions on Knowledge and Data Engineering, vol. 16, No. 10, Oct. 2004; pp. 1-17. |
| D. Saputra, et al.; Mining Sequential Patterns Using I-Prefixspan; Proceedings of World Academy of Science, Engineering and Technology, vol. 26, Dec. 2007, ISSN 1307-6884, pp. 499-504. |
| Y. Xu, et al.; Analysis of Web Access Sequence Based on the Improved Prefixspan Algorithm; International Industrial Informatics and Computer Engineering Conference (IIICEC 2015), Atlantis Press; pp. 788-791. |
| Pegasystems Inc.; https://www1.pega.com/products/pega-platform/robotic-automation; accessed Jan. 9, 2019. |
| B. Dalager, et al.; Changing the Future With Robotics; 44 pages; Pegaworld 2016. |
| G. Arora, et al.; The Emerging World of Humans & Intelligent Robotics: Working Together to Build Smarter Enterprises; 29 pages; Pegaworld 2016; Pegasystems Inc. |
| Pega Academy; Pega Robotic Automation Architect Essentials; 1 page; Nov. 2016; Pegasystems Inc. |
| Pega and Marketforce; The Future of Work; 24 pages; 2017. |
| A Pega Datasheet (Pegasystems, Inc.); Pega Robotic Process Automation; 2 pages; 2018 Pegasystems Inc. |
| S. Merritt; Building and Scaling Robots that Deliver Impact: The Leap From RPA to Digital Process Automation; 13 pages; 2018 Pegasystems Inc. |
| Pega Infinity; Pega 101: Connecting Digital Customer Engagement to the Digital Process Automation that Makes it Happen; 53 pages; 2018 Pegasystems Inc. |
| Pega Platform; Pega Robotic Automation—Using the Automation Recording Feature 8.0; Jun. 21, 2018; 16 pages; Pegasystems Inc., Cambridge, MA. |
| Pega Platform; Pega Robotic Automation—Using Data Components 8.0; Jun. 21, 2018; 30 pages; Pegasystems Inc., Cambridge, MA. |
| Pega Platform; Pega Robotic Automation Studio—Creating UI Connectors 8.0 and 8.0 SP1; Jul. 30, 2018; 30 pages; Pegasystems Inc., Cambridge, MA. |
| Pega Platform; Pega Robotic Automation Runtime—Installation Instructions 8.0; Aug. 21, 2018; 24 pages; Pegasystems Inc., Cambridge, MA. |
| Kryon Systems; https://www.krvonsystems.com/process-discovery/ ; accessed Jan. 23, 2019. |
| Pega; Leveraging Robots for the Customer Journey; undated; 27 pages. |
| Number | Date | Country | |
|---|---|---|---|
| 20200219033 A1 | Jul 2020 | US |
