Robotic Process Automation Archives

G. N. Shah August 16, 2022 No Comments

Is Your Company Ready For Process Automation?

What are the factors that suggest your company (or at least a specific subset of processes within your company) are prime candidates for process automation? I’ve been asked that question about process automation more often than any other by company executives. That is especially the case for companies in the midcap market arena, where process automation has not yet penetrated significantly. These executives hear the hype about process automation but don’t really know if they are a good candidate for implementation. They do not want to devote their time, resources, or management focus to pursuing this innovation if it is not a good fit. Some proponents of process automation say all companies should be pursuing this innovation, and all other things being equal, I agree that is the case. But the reality is – again, especially in the midcap market arena – that is not always the case. Companies may need to pursue many other even higher priority initiatives to improve their company and advance growth. So, how can a company decide if they are well suited for process automation, and now is the time to pursue it? In this short note, I lay out 7 factors that I feel are strong indicators that implementing process automation will have significant positive impacts on your company, including the economic benefits but going well beyond that. [1] Are these the only 7 factors? No, there are many other factors and predictors of success, many of which are outside the technical scope of the process automation project. For example, having and keeping strong executive support on the project throughout the project implementation stage and continued support during the ongoing operations and expansion stages. These types of factors are not the ones I concentrate on in this note. The factors I point out here are ones specifically related to the process areas of your company to be automated and linked to why automation of those processes is likely to lead to big benefits for your company. Microsoft Power Platform A quick side note before we begin. My company, Innovatix Technology Partner, is a Microsoft Gold partner and our practice in process automation revolves almost entirely around the Microsoft Power Platform suite of tools.[2] We have large teams trained and ready to go to implement your process automation needs using this suite of tools. Depending on the work at hand, we use all 4 of the major tools in the suite as well as the 2 new emerging tools. The 4 main tools include Power Automate, Power Apps, Power BI, and Power Virtual Agents. The two new emerging tools are Process Advisor and Power Pages. And of course, Microsoft is devoting major resources to further advancing their power platform tools set, so more and even better capabilities can be expected going forward.[3] Innovatix’s company website gives detailed information on all these tools and some guidance on the conditions when one is best to use, depending on a company’s needs and goals. Some factors include the desire to pursue low or no-code implementations, the need for process mining to uncover outlier processes, the need to publish output information to web pages, etc. regularly. So, in my discussion below of the 7 factors that favor process automation, I am thinking of these factors in the context of building and implementing the automation applications using the Power Platform suite of tools. The six tools in the suite are highly integrated and thus allow us to weave them together as needed and appropriate in the automation applications we build. While I am sure other process automation tools in the marketplace must have similar capacities, I believe the highly integrated nature of these 6 Microsoft tools is indeed a huge and perhaps unique value addition. For example, imagine the benefits of being able to output process automation monitoring statistics via Power BI. Now, on to the 7 factors…. Factor 1: Lots of manual operations This is considered the perfect case for process automation. The classic case is cited in the literature repeatedly for implementing process automation. I am not talking here of sales or call center operations, for example, which certainly have and need many employees to perform the necessary functions.[4] Instead, I am stating here the areas of your business where you have a lot and growing number of employees doing significant manual work, including the following use cases: If a lot of people in the company are spending a good amount of their time doing these manual operations, then that provides the basis for taking a good look at process automation. Especially in the case where you consider these manual operations as a growing part of your operations. There is a lot to dissect in each of the 4 use cases listed above. Each of these four cases I cite above[5] has specific signatures for how best to automate. I will not discuss each of these signatures in detail in this note but will return and do that in one of my next notes. Let me just give a flavor of what that discussion will look like. The first use case above can be handled in a bunch of ways. But one clear simple way is to just do the automation by building new APIs or adapting existing APIs in the two systems that make it possible for the required information to be transferred automatically from one system to the other.[6] This is a straightforward case of process automation. Microsoft Power Automate will provide the cloud administration platform for managing and monitoring these new automated process steps. Factor 2: Outlier processes I call this factor the ‘hidden outlier process’ factor. Namely, you have business processes that should be well-defined based on your business process mapping blueprints, but unfortunately, on many days, some bottlenecks and fall downs cause manual intervention to take hold of various processes, activities, and customer orders. These ‘outlier processes’ are unexpected (at least in their intensity and frequency) and seem to be causing much

Ronald Mueller November 14, 2021 No Comments

How Machine Learning & AI Are Transforming The Future

Machine Learning (ML) and Artificial Intelligence (AI) are like two sides of the same coin, with artificial intelligence being the underlying technology driving machine learning. Machine learning delivers a huge amount of data, which can be processed using big data analytics and appropriate predictions can be derived upon using AI technology. This is a next step towards transforming the future as human intervention is no longer a necessity. It is possible to setup an error-proof system with this sort of technology, ensuring greater efficiency, productivity and high levels of accuracy. Traditional legacy IT systems aren’t flexible enough to quickly scale with the ever changing digital economy. Need for greater efficiency and faster decision-making are driving enterprises to bank on modern technologies leveraging a combination of big data analytics, artificial intelligence and mechanical automation. Machine learning and AI form an inevitable force driving innovation across information technology, effectively serving three specific segments: Customer Service and Retention Management – Bots also known as web robots have gone a long way in bringing about greater customer satisfaction. These automated applications perform highly repetitive tasks having the ability to process natural language like a chatterbot program simulating human talk. Bots are currently being used in automatic railway enquiry, banking services, hotel reservations, shopping guides or even scheduling appointments. Such services heighten customer experience resulting in loyalty and most of all retaining the customer base. Business Operations and Process Management – Automation alone can greatly improve the efficiency of business processes and workflows. Redundant processes that involves unnecessary human intervention are taken care off during automation reducing operational overheads. This when combined with AI enables organizations improve operational snags with real-time analytics and extended intelligence being retrofitted onto the system. AI makes sense of the vast volume of redundant data by forecasting useful statics that helps plug process inefficiencies. Larger organizations are focusing their research in AI to reduce maintenance costs by anticipating breakdowns and operational failures. Cyber Security and Risk Management – Ever increasing cyber risks have to a great extent propelled AI research. AI’s ability to quickly track patterns and deviations in vast amounts of data gives it a sort of sixth sense in tracking potential threats. With machine learning around the corner, threats can be continuously monitored in real-time averting any security breaches. This system is already being used in banks to prevent unauthorized access of user credentials and in the IT sector safeguarding business critical data from networking threats and hacking. One obvious question that arises is – are we losing control with AI and machine learning. As always, any new technology comes with pros and cons, as long as they are deployed within the business framework. Same holds well in case of AI and machine learning, with AI suggesting on how to improve operations and mechanical systems learning to generate their own algorithms. In spite of all this the human intervention will play a judicious role in deciding whether to implement such recommendations or forgo it. Take for example, IBM’s supercomputer Watson that’s predicting patient’s condition with a reasonably good accuracy by way of continuous learning. Similarly, IT barons like Amazon, Google, Flickr and Netflix are changing their digital business models by incorporating advanced automation and AI techniques. We are yet to witness a mechanical revolution powered by AI taking over the modern era. Advanced AI technologies coupled with robotics will find better ways of delivering businesses across the globe with least human intervention. Business organizations can pin their hope on future machine learning models to deliver a shared architecture across businesses. As of now we have only explored the tip of the ice berg with more to emerge. Let’s hope AI will pave the path for a high-tech automated future.

Rajaganapathy K July 22, 2021 No Comments

The Microsoft Power Platform Tool Set: A ‘Down-to-Earth’ Primer

This paper provides an overview and practical guide to the tools within Microsoft’s Power Platform family. Among other things, this set of tools is intended to provide automation capabilities to users. That is, the tools provide customers with low-code/no-code tools for automating business functions and processes, including reporting and dashboarding. The tools are all state of the art, and well regarded by technology rating companies such as Gartner, with some being clear leaders in their sector. Our expectation is the tools will all widen their lead in the next few years. No surprise, all tools in the platform work seamlessly together and work well with other products in the Microsoft family, including Office 365. They also work well with hundreds of other enterprise apps, with Microsoft already having built the connectors needed to pass information back and forth between these apps. The tools in the Power Platform are all reasonably priced, in most cases well below principal competitors, and are easy to set up and use quickly, especially for Microsoft-based companies. We are writing this paper for two reasons: The net result of a company adopting these tools will likely include: many robots automating multiple mundane processes and tasks; much greater accuracy in data flows throughout the company; much greater concentration of human work on value-added and customer-oriented tasks; and of course, a major positive boost to ROI normally within-year. For the past several years, the hype for both RPA and digital transformation has been at a fever pitch. What we are hoping to do here (and in all the other papers and posts on RPA on our website) is brush aside the hype and provide practical, down-to-earth information on each of the tools and how they can work together to solve practical business problems and how they can be used to automate straight-forward business processes within mid-sized companies. In other words, they work, they are relatively easy to use, and once in place, the robots are robust and secure and do indeed achieve the promised world of robotic process automation! As noted earlier, our focus in this paper and our professional services is targeted to mid-sized companies (or to a mid-sized organization within larger companies). Leaders in those companies know they must act soon and forcefully to automate their operations but may have trouble seeing clearly through all the din of all industry and marketing hype. Our clear recommendation to you is to engage the Microsoft Power Platform set of tools to get the job done. Contact us anytime so we can explain our point of view further. Microsoft Power Platform The chart below shows the tools grouped under the Microsoft Power platform. They include: As also show, three general capabilities are common across all 4 tools, namely: hundreds of already-built Data connectors, an easy-to-use AI-builder that can be invoked in the same way across the 4 products, and a common Dataverse for securely storing and managing data across the 4 apps. The Power Platform set of tools enables a customer to: Commonality and synergy are the key themes behind these four tools in the Microsoft Power Platform family. Users can take advantage of common data connectors to easily pass data and information between the different tools. They use common capabilities and functions, which greatly improve productivity and accuracy in implementing the tools in a company’s environment. We honestly believe a mid-range Microsoft-oriented company should look no further than this family to power up their use of robots and gain the advantages of automation, including in reporting and dashboarding. Power Platform Family of Tools In this section, we provide short summaries of each of the four tools. Our point of view in these summaries is to keep our eye on the Power Automation tool and describe how the other three tools and the three general capabilities listed above can play significant roles in augmenting the Power Automate tool. Power BI – Used to analyze data from different data sources: It is the premier self-service business analytics tool provided by Microsoft. Power BI has a few offerings starting from the Desktop version to the Power BI Service, hosted on the cloud. It can connect (via pre-built connectors) to a wide range of data and apps. The tool can be used to design interactive reports, dashboards, or stories, all supported by compelling visualizations. From a Power Automate point of view, we see Power BI as the reporting tool for tracking and spotlighting all the KPIs used to oversee the success of the various robots introduced via Power Automate. So, log data generated by the Power Automate robots gets directed to the Dataverse and then can be used by Power BI to show all the relevant tracking results. Power Apps – Used to build powerful mobile apps for internal use by an organization: It is an intuitive platform that provides users with drag and drop features to build a user interface for a mobile application. The user can add various controls to the user interface, including textboxes, choice fields, etc. It also allows users to add in media devices like the camera, videos, etc., and other related features necessary to build a modern mobile application. There is a feature to connect to various data sources using Power Apps, and after the development is completed, a user just needs to publish the app for it to be available within the organization. Here we see the connection to Power Automate as follows. Once multiple robots (often 10-20-30) have been built and running within the organization, Power App can be used to build a tracking and alerting app to provide the automation team with real-time feedback on how well things are running and if anything needs to be investigated. Power Virtual Agents – Used to develop flexible chatbots that can communicate with internal staff as well as external customers: This is a new addition to the Microsoft Power Platform. Power Virtual Agents is the bot-building service provided by Microsoft for business users. Using this, a user

Rajaganapathy K July 8, 2021 No Comments

Automation of An External Data Feed Extraction and Assembly Process

This is the second in a series of posts summarizing case studies on robotic process automation (RPA). These posts describe a growing number of Innovatix Technology Partners (formerly Macrosoft, Inc.) projects done recently to solve client problems as well as several to automate internal processes within our company. These are real-life case studies of RPA. The automation apps we have built and tested are now working as intended or are ready to move into production. In all cases, we are using Microsoft Power Automate to build and deploy the automation robots. We are doing this post series to show the varied nature of RPA apps, from ultra-simple to highly complex over a broad range of business processes. We hope readers will find parallels in their own business processes. Stated simply, our view is that, that companies should seek out all processes that can be automated and build RPA apps to do that work. Gartner defines this view as ‘hyper-automation, namely, “the idea that anything that can be automated in an organization should be automated.” We hope this series of RPA case study posts will stimulate readers’ further interest in moving their companies in that direction. We are here to help. Case Study 2: The Automation Challenge This case study involves the automation of external data extraction and assembly process for one of our clients. Innovatix has been providing managed services to this client for many years, including developing and supporting several internal apps and their public website. We have also been giving integration services for their website to other enterprise apps, including Salesforce CRM and an external email marketing and delivery service. We set about trying to find mundane tasks to automate within this company, and in conjunction with the client team, we came upon the current external data assembly process. The business requires a complex data extraction and assembly process involving gathering data from multiple external FTP sites, websites, and emails daily. The data extraction and assembly process was being done entirely manually until we undertook to automate it. Here are some of the challenges the client faced with the current manual process. These had to be addressed in moving to a new automated process. Multiple Data Providers, Data Sources, Formats: Pulling source data was complicated and time-consuming. The external sources of the data had different formats, structures, and types. Moreover, once the data was fetched, it needed to be validated to make it compatible with the destination system before making it available for internal and end-users. Data Integrity, Transformation & Security: Data quality was a primary concern in handling the data from multiple external sources. Poor data quality would create a compounding problem that could lead to compliance issues. If invalid or incorrect data were passed downstream in the data processing stages, it could lead to corrupt and incorrect results, a major problem for the company. Scalability: The client is faced with significant heterogeneity of the data from the diverse sources and the need to assemble and standardize all this data into a unified data structure and system. The client is expecting to add many more data sources in the near future, which will lead to substantial growth in data volume over the current situation. To tackle this challenge, the client required us to employ a robust integration solution that can handle high volume and disparity in data without compromising on performance. Deeper Dive into the Case Study As noted above, this second case study involves a client wanting to automate and synchronize data extracts from a whole range of external sources. We automated this process using Microsoft Power Automate, our go-to automation tool[1]. There were three parts to the automation as outlined below. Each was done in succession using Microsoft Power Automate. All three automation stages are done, and the robots are in production handling the data flows. Stage 1. FTP Sites: The client connects to many FTP accounts of vendors to fetch data made available by these vendors. The first critical issue in automating this stage is security – security of access and connectivity to these data repositories is critical. The client requested a cost-effective, efficient, and scalable solution for automating the file transfer from multiple external FTP servers to the client’s internal enterprise Microsoft OneDrive platform. Manual data transfer processes are vulnerable and subject to human error, making them inefficient and often unreliable. In automating these data transfers, our requirement was to reduce or eliminate the need for manual file exchanges and ensure high standards of data security. Automating these file transfers takes the manual file exchange burden off individual team members. Innovatix implemented this process using Microsoft Power Automate SFTP connector for automating the file import process and validating files received from each of the data providers. The automation solution was set up to alert designated parties with notification emails whenever the system encountered a delayed or missing file from any one of the data suppliers. Also, the automation process significantly reduced manual verification efforts and, at the same time, restricted access of employees to multiple FTP servers. Stage 2. KPI Index Sites: The client then requested Innovatix to automate collecting and aggregating multiple KPI files and assembling the individual KPI index files into one file format (pipe delimited). Automating opening the individual KPI index files from the various providers was the key to this stage of work. The aggregated pipe-delimited KPI index files needed to be saved to the client’s OneDrive platform, and a copy had to be uploaded to an external vendor. The aggregated file included two additional columns for the date & source of the data generated dynamically when each file was received. Other fields in the data had to be updated at specific times of the day to reflect conditions at those times, instead of the values in place when the files were extracted from the vendor site. Innovatix has demonstrated the automation capability of Microsoft Power Automate in the prior stage of the project by creating automated workflows

Ronald Mueller June 17, 2021 No Comments

Microsoft Power Automate Desktop: Capabilities and Instructive Scenarios on How to Use It

This paper is about Microsoft Power Automate Desktop. We describe the major capabilities in the tool and then show how these types of capabilities can be used in different types of process automation scenarios. So, think of this paper as a simple learning primer on Power Automate Desktop for readers who have not yet worked with it or more generally have not yet engaged with robotic process automation. We hope you will find this paper a down-to-earth guide to when where and how to use Power Automate Desktop to start moving yourself and your company in the direction of complete digital automation. As a Microsoft Gold Partner, we are ramping up our efforts across the entire set of tools in the Microsoft Power Platform, and we will be publishing many new papers to highlight this area of our company’s expertise. We are investing heavily in the training and certification of our developers and data engineers in the full set of tools, but most especially Power Automate, of which Power Automate Desktop is one component. We have other papers in this series describing Innovatix’s initiatives in this area, including a series of case studies of some of our recent projects built using Microsoft Power Automate. One final note. Innovatix’s commitment to RPA is consistent with the view introduced by Gartner in one of its recent analyses of the state of technology for RPA. Gartner defines this view as hyper-automation, namely “the idea that anything that can be automated in an organization should be automated”. Introduction In this white paper, we discuss the major capabilities offered by Microsoft Power Automate Desktop. Microsoft Power Automate Desktop (PAD) is a workflow automation platform that helps automate rules-based mundane tasks on the desktop or the web. It offers a tremendous and growing set of functionalities including conditionals, variables, OCR automation, desktop automation, web automation, and more. We will discuss each area and shed light upon some of the functionalities offered in each area that might spark your interest in the product. PAD has more than 360 prebuilt actions, so we are not going to go through all of them. After discussing the capabilities of PAD, we will dive into some scenarios where PAD might help you in the daily tasks that you perform on your computer. We expect to be adding more such scenarios in the coming months, as we encounter new and interesting ones. Our goal is to highlight real-life workflow examples where PAD can be very helpful so readers can see how easy and quick it is to get this same case study going on their own desktop or the web. PAD Capabilities Here is a starter list of 13 capabilities in PAD and summary descriptions of each: Capability Description Where Useful UI Automation PAD lets you click a UI element in a window, select a tab in a window, select menu options, drag and drop UI elements in a window and expand or collapse a tree node in a window. PAD also offers us form filling functionalities including populating a text field in a window, pressing a button, selecting a radio button, setting a checkbox state, and setting up drop-down list values in a window. These features are helpful when we encounter automation where we need form-filling capabilities and need to automate applications on the desktop. Web Automation PAD allows us to launch browsers including Chrome, Firefox, Edge, and Internet Explorer. Create a new tab in a browser, click a link on a web page and close a web page. Extract data from a webpage and take a screenshot of a web page. These features are useful when making automation that requires navigating through a webpage and extracting data from them. After extracting the data, we can paste it into files if this is the required use case. Excel Launch new or existing excel files. It also has read and write capabilities. When we combine PAD Excel with PAD Desktop it creates a powerful result as it allows us to use various excel functionalities through clicking and selecting the radio buttons and more. Email and Outlook PAD allows us to retrieve email messages, process email messages, and send email messages. We can also launch Outlook and retrieve, send, process, and save email messages in the Outlook application. This is useful in all alert use cases as well as applications where data and information need to be communicated to others. Mouse and Keyboard We can get the mouse position, move the mouse to an image or text, send a mouse click and send keys. PAD allows us to use all the Keyboard keys including the special keys. This is helpful in automation where we need to fill in forms on web pages for example signing in into an account. Conditionals and Loops PAD offers a plethora of conditionals besides the basic else, including else if, if and switch for example if file exists, if folder exists, if process exists, and more. Then PAD offers loops that include each loop. These advanced conditionals help in use cases where basic conditionals fail. Each loop is useful in iterating through a list and performing a bunch of actions repeatedly. Wait PAD offers a bunch of wait functions which include wait for file, wait for a process, and more. Sometimes when making automation the next process starts before the previous one ends. This causes the automation to fail. In these cases, we use wait functions which adds a delay in the start of the next step and therefore gives time for the previous step to finish. Variables Variable functions are used to store data for further processing in automation. PAD has a diverse range of variable functions which include generating a random number, truncating a number, clearing a list, merging lists, and more. System PAD System functions include taking a screenshot, emptying the recycle bin, locking the workstation, logging off the user, running PowerShell scripts, print documents, and more. System functions aid in building automation

Ronald Mueller January 12, 2021 No Comments

Analytics & RPA: The Transformative Digital Enabler

Analytics & RPA is the Key to Major Success RPA is currently one of the most widely successful digital technologies. RPA enables companies to reduce human errors, operate at much higher efficiencies, and increase employee satisfaction by shifting focus from repetitive menial tasks to value-added roles, and in the process further enhancing customer satisfaction and innovation. Companies are now going one step further – towards Intelligent RPA. In many automation journeys, cognitive technologies such as AI, ML, NLP when blended with RPA, bring smart capabilities to existing RPA initiatives and help magnify and enhance the positive outcomes vis-à-vis standard RPA. Analytics is essential to providing RPA users with the measuring tools and KPI visibility necessary to ensure they maximize the value of their automation and understand the quantitative impact of these automation efforts on positive business outcomes. In this blog, we discuss one analytics framework that opens the door to a full understanding of your RPA program – UiPath Insights. If you would like to learn more about UiPath Insights, or about analytics in general in the context of RPA, please contact us, we will be happy to discuss further. UiPath Insights is THE Analytics platform for RPA UiPath Insights is an end-to-end RPA analytics solution that enables users to track, measure and forecast the performance of their entire automation program — so users know the impact it’s having and how to scale to the next level. The basic steps needed to get UiPath Insights up and running include: The chart below shows the basic steps in UiPath Insights implementation, and also shows what you gain for your company once it is in place. With UiPath Insights you can measure the true impacts of RPA on your Processes and your Business Among the most important benefits of UiPath Insights are the following four. The rest of this section briefly describes each benefit and its impact on your RPA automation initiatives. 1. Start measuring so you can start improving Measure the performance of your RPA operations, including situational awareness of events happening in your operations. You can measure and analyze everything from robot and process performance to individual transactions and exceptions. Start with the robot and process dashboards to get immediate visibility into your RPA operations. See the example below showing a robot performance dashboard. 2. Calculate RPA’s Impact on Bottom Line Each process in your operations requires a unique set of KPI’s to define success. With UiPath Insights you can customize the way you calculate and align your automation goals to the distinct needs of your business. Start by tracking the time and money saved by each of the automation you have implemented. You can generally do this dashboard right out of the box, with little customization required. See the example of business ROI shown in the chart below. 3. Forecasting and Anomaly Detection Utilize historical data to forecast future operational states, and ML-based anomaly detection and alerting to get notified about critical events and major milestones. Easily gain new analytical perspectives of your operations with Insights-recommended breakdowns of your existing data. Start by forecasting robot utilization to ensure you are maximizing the benefits of your RPA robots in all your processes. See the example time chart below. 4. Shareable Reports The value of performance reporting should continue past a set dashboard – it should drive strategic discussions and action across the entire enterprise. With UiPath Insights, you can easily share reports with key stakeholders and process owners. Start by creating a drag-and-drop report based on one of the existing templates provided in the system, and then set it to be shared with the entire team on a regular schedule. Conclusion: Analytics can significantly boost your RPA advantage Analytics is key to measuring and assuring that your RPA automation initiatives yield the best possible results for your company. Contact us to discuss further how we can assist you in implementing an analytics framework, like UiPath Insights, in conjunction with your RPA program.

G. N. Shah December 14, 2020 No Comments

Incorporating NLP Capabilities into Innovatix’s WebMineR

Executive Summary Innovatix Technology Partner’s new product, WebMineR, is a state-of-the-art web scraping cloud application. It has over 25 best-in-class capabilities, and is highly scalable, efficient, secure and configurable. The list of major features continues to grow, as shown in the latest Roadmap posted on our web site. In addition, several recent papers on Innovatix’s web site discuss in detail these 25 major capabilities, and how they compare to other products in the marketplace. This paper starts a dialogue on some of the newest functionality to WebMinerR which are centered in the area of NLP, specifically text summarization and topic modeling. These are clearly important features to have in a web scraping technology. We expect these capabilities to be fully available by second quarter 2021. The purpose of this paper is to describe the research into NLP we are now doing to make sure these capabilities perform well when delivered in WebMineR. Text summarization and topic modeling are two of the most prominent use cases in the field of Natural Language Processing (NLP). Text summarization allows one to understand the basic idea of a block of text without having to manually read and summarize the document. Topic modeling on the other hand provides the ability to extract main topics from a block of text and thus get the key ideas of the paper, again without manually reading it. That is at least the concept behind these two important NLP capabilities. Certainly, the technologies are getting better with time, as AI/NLP continues to move forward at a tremendous pace. How well they actually work today for text extracted from a wide range of web sites (in many different languages) is still an open issue in our minds. For sure it is critical to know how to best configure and use the NLP algorithms we select to use in these two areas in order to get the best possible out of them. Hence, that is the goal of our current research in this area and below we report on some of our findings to date. We are doing this testing and research in order to ensure they are reliable when we incorporate them into our web scraping system. WebMineR is built to scrape public information off web sites at ultra-high throughput capacity. The addition of text summarization and topic modeling to the tail end of the WebMineR process would be an obvious major benefit to our user community. This paper contains a detailed background on both these subjects – text summarization and topic modeling – along with some results of trials and case studies of different off-the-shelf NLP systems and algorithms we have tried out so far. We expect to continue this testing and research through the end of the year. The tests we show here are on healthcare-related documents. In addition to English, we show how these capabilities can be applied to Mandarin language documents as well. Overall, our research and testing results show reasonably good performance using NLP methods to extract summaries and topics from documents. We feel strongly this will add useful new capabilities to WebMineR. Stay tuned! Text Summarization Text summarization is the task of creating a concise and accurate summary that represents the most critical information and has the basic meaning of a longer document’s content. It can reduce reading time, accelerate the process of researching for information, and increase dramatically the amount of information found within an allotted time frame. Text summarization can be used for various purposes such as medical cases, financial research, and social media marketing. Text summarization can broadly be divided into two categories: Extractive summarization and abstractive summarization. Resources/libraries that are commonly used for text summarization include: Natural Language Toolkit (NLTK) and Spacy, and pre-trained models like Bert and T5. NLTK and Spacy are open source libraries used for NLP on Python. Bidirectional Encoder Representations from Transformers (BERT) is a technique for natural language processing pre-training developed by Google. BERT is pre-trained on a large corpus of un-labelled text, including the entire Wikipedia(2,500 million words) and Book Corpus (800 million words). You can fine-tune it further by adding just a couple of additional output layers to create state-of-the-art models for your specific text mining application. T5 is an encoder-decoder model and converts all NLP problems into a text-to-text format, and it is trained on a mixture of unlabeled text (C4 collection of English web text). In our test use cases, we used NLTK, Spacy, and BERT for extractive summarization, and we used T5 for abstractive summarization. We applied each of the 4 tools to implement text summarization. Here is an example that shows the summaries that we generated using the different tools. We first load a text file (PDF format) and then applied the four text summarization methods. Text summarization by using NLTK library first uses Glove to extract words embedding and then uses cosine similarity to compute the similarity between sentences and apply the PageRank algorithm to get the score for each sentence. Based on the sentence scores it put together the top-n sentences as a summary. Spacy library will tokenize the text, extract keywords, and then calculate the score of each sentence based on keyword appearance. Text summarization using pre-trained model BERT will first embed the sentences and then run the clustering algorithm to finally find the sentences closest to the centroids and use those sentences as the summary. T5 converts all NLP problems into a text-to-text format, and the summarization is treated as a text-to-text problem. The model we used is called T5ForConditionalGeneration, and it loads the T5 pre-trained model to extract a summary. The document we used to implement text summarization is a summary of the European public assessment report (EPAR) for the drug Fosavance. It explains how the Committee for Medicinal Products for Human Use (CHMP) assessed the medicine to reach its opinion in favor of granting a marketing authorization and its recommendations on the conditions of use for Fosavance. After analyzing the 4 summaries, the

Imaan Ali December 8, 2020 No Comments

Common Causes of Failure Among Process Mining Initiatives, and How to Avoid Them

Executive Summary Process Mining is an extremely useful tool in the practice of modern process automation. It not only helps to ensure processes and data align with company expectations but also can support organizations in maximizing process efficiencies and in helping companies focus on ways to improve customer service. Despite the usefulness and applicability of process mining in most business processes, there are several major pitfalls that can severely handicap how well process mining works if not addressed properly, and the best time to do that is before implementation begins. In this blog we identify and describe five such challenges that new users of process mining need to make sure to avoid. Keep in mind, process mining encompasses a tremendously growing and expanding field of automation techniques, so for customers considering use of process mining for the first time there is an ever-expanding set of features and techniques from which to choose. Of course, the critical consideration is which ones are the most beneficial for your company right now and in the immediate future. With the rate of change in the tool sets, it is unrealistic to factor in projections of the long term when it comes to process mining. As noted, process mining tools are constantly under development and have ever expanding sets of major features and capabilities, including, most importantly, ever-increasing amounts of AI modelling and projections throughout the entire process mining framework. Because of this, it is critical that companies not be entirely dependent on the professional staff or partner consultants for the selected process mining tool. Company personnel need to be key contributors to the implementation of process mining on all a company’s processes. Companies need to take the significant step necessary to educate internal staff on the selected tool (and on process mining in general), to ensure that the tool succeeds and provides significant benefits to the company. Here are some of the most common challenges companies face in their process mining initiatives. Since process mining tools are constantly being updated and new features and capabilities are being added, these particular challenges may lessen in importance over time and/or be replaced by new challenges due to advances in the tools. Using Process Mining for a Process that Doesn’t Really Need It While choosing the tool most appropriate for your company is a critical step, it is also important to analyze your own company’s processes thoroughly and take enough time to figure out where process mining can be best used and provide the most benefit. There are several signs in a process that can make it a very good candidate for process mining. These include overly complex data; no clear data structure apparent; unidentifiable problems, deviations and bottlenecks found in the process- just to name a few. By picking a process that is not entirely in need of process mining, a company is essentially wasting its money for no important reason. The way to ensure that a process mining tool will be maximizing company benefits is to make sure to deploy it for a process that checks off the most boxes of being a good candidate. Although using a process mining tool does not automatically correlate to maximizing a company’s profit, picking the most appropriate process to use the tool to start with will definitely increase a company’s chances at success and will greatly increase the internal momentum within the company to continue to expand the processes covered by process mining. The Process Contains Incomplete Data Having incomplete data in a process will be an obvious bump in the road towards achieving full process automation. Since process mining tools are specifically designed to find gaps in the data, using process mining for a process with incomplete data can make some sense in helping process engineers to better understand these gaps. However, it will definitely be easier and more straightforward if these gaps can be filled before the process mining tool is implemented. Depending on which part of a process the data is missing, if there is missing data, the resulting process mining outputs may be significantly degraded as process mining tools produce a fully linked chain of events where each successive event impacts the rest of the actions. If data is missing in one part of a process, that will impede progress in gaining a full understanding of the process using a process mining tool. The Process Has Too Much “Concept Drifting” “Concept Drift” refers to the tendency of processes to change over time, in particular while they are under the ‘analysis microscope’ of the process mining tool. For example, process changes can occur during different seasons; this is an example of how processes can change periodically. The chart below demonstrates the various types of concept drifting that can occur in a process, whatever the underlying reason may be. These drifts can impact the digital structure that process mining comes up with and will make the process mining analysis and results very unstable and difficult to interpret. Splitting a process up into several smaller event logs is often an easy and effective way to discover possible concept drifting. Being aware of and managing concept drift in your process from the beginning before process mining is first implemented will help ensure that it remains controllable and understandable as process mining procedures are implemented, and prevent ‘drifts’ from having negative impacts on process mining analysis and results. Data is Not in an Appropriate Format Pre-processing and data preparation are vital steps to a successful process mining initiative and unfortunately this step is often overlooked. It is crucial to look at the import format of your data to ensure that it agrees with the process mining tool you have selected. If there is a mismatch in format, the data will likely not be processed fully or even at all. In these cases, analyses from the process mining tools will be of poor quality, showing results with missing spots or even a completely blank event log. Clearly this needs

Imaan Ali December 3, 2020 No Comments

Document Understanding: A Short Guide on the Major Process Steps

1. Executive Summary Document Understanding (DU) is one of the fastest-growing areas in business process automation. The DU ecosystem includes technologies that can interpret and extract text and meaning from a wide range of document types including structured, semi-structured and unstructured — even ones that contain handwriting, tables and checkboxes. This is now possible because of the ever-improving techniques of Machine Learning (ML). Enhancements in ML are spurring innovation in document understanding. In this article, we present the major steps in the DU process and the underlying architecture with reference specifically to UiPath’s Document Understanding framework. Multiple technologies can unlock the power of document understanding such as: In today’s business processes, most of the routine and mundane tasks employees perform consist of creating, reading, reviewing, and transcribing paperwork (documents). Employees spend a significant percentage of their work time reading these docs, extracting data, and passing on the much-needed information into other downstream applications manually. Since the data extraction from the documents and input to other apps is done by a human, the process is subject to problems of accuracy and reliability. UiPath’s Document Understanding solution allows you to intelligently process data with a high level of accuracy and reliability for any type of document such as invoice, receipt, financial statement, utility bill, and any other kind of text that has a different structure. The general flow for UiPath’s DU process is encapsulated into the 6 process steps below. To decide which steps are needed for a specific business process, you will need to address the below requirements: One primary concern of the solution is that it should not stop the entire process until a human performs a manual verification. The process should escalate the check to the respective party, while at the same time continuing to evaluate and process the rest of the documents. 2. Classification Based Approach There are scenarios where data extraction is not essential, and the priority is only to segregate the documents based only on classification, for further processing later in another process. In such cases, the UiPath DU solution comes in very handy as it provides the capability to classify documents based on keywords. The solution offers the ability to train the classifiers intelligently when setting up the automation solution. These classifiers will also continue to learn every time a document is classified (and verified by a human) thus improving accuracy over time. The classification and verification process steps are suitable for attended automation. The attended automation provides a Classification Station, where a user can verify and correct the classification if the confidence is below a predefined value. A schematic of the process is shown in the below chart. Classification Process In most business scenarios, classification is not the only requirement. Most processes will also require the extraction of data from the documents and processing of the extracted data according to specified business requirements. However, even in this case, we cannot ignore the classification process step in the automated approach, as it is essential to identify the type of document so that the robot knows how and what fields to extract. Different methods are available to handle manual verification of classification results: Use attended and unattended collaboration in scenarios where the process should be manually triggered. If the same user who triggered the process is doing the validation, the use of a Validation Station is possible. However, based on the business logic, if certain exceptional cases need management approval, such escalations can be directed to the Action Centre directly without showing to the user at the Validation Station. When designing a Document Understanding solution, it is a good practice to break the solution into separate manageable sub-processes. As a generic solution that fits for most cases, we could introduce three sub-processes to handle the Document Understanding Framework. This high-level diagram showcases a sample architecture for the Document Understanding process. The architecture used here breaks the entire document understanding process into three main sub-processes. The three main components are Initiator process along with processing logic (Process 1), UiPath Action Centre for task assignment and management (Process 2), and finally, the Train models component (Process 3) which handles the training of the intelligent classifiers and also the passing of the extracted data to other applications. The detailed architecture of each part is as follows. Process-1: The initiator Process The Initiator process is the primary process that handles document classification, data extraction, and verification logic. The verification logic will include the rules that define how to handle verifications automatically, either through the use of the Validation Station, Action Centre, or both when human intervention is needed. Depending on the option chosen in the validation logic, the extracted data will finally be passed to either the Action Centre Processor or to Post Processing to continue to the next steps. The diagram below shows a sample architecture for the Initiator Process. Process 1: Initiator Process Process 2: Action Center The Action Centre is the process that handles task creation, waits for task completion, and finally passes the data to the Post Processing portion for the end of Document Understanding. The diagram below shows a sample architecture for the Action Center process. Process 2: Action Center Process Architecture Process 3: Post Processing The post-processing includes the steps needed for exporting the final verified data, training the models, and finally, passing the data to a different process outside the document understanding framework to continue with any system interactions, etc. The data is handed over to a separate process because such steps are not part of document understanding and those should be maintained independently to maintain integrity and reusability of each component/ process. The diagram below shows a sample architecture design for the final stage of the document understanding framework, and it also showcases how Process 2 connects with Process 3. Process 3: Post Processing Solution Architecture Conclusion Although the specifics of the business process may change from one company to another, the core architecture and process steps showcased above remain largely the same

Imaan Ali November 24, 2020 No Comments

Is Your Business In Need of Process Mining?

5 Signs that Your Business Process is too Complicated and In Need of Process Mining As more automation opportunities arise, it is becoming increasingly imperative for businesses to use process mining to examine their business processes closely and determine those most ready to yield significant benefits from automation (and not solely focus on the latest technology stack!). Getting a more complete view of reality of a company’s processes and understand precisely what is happening there in real time and at the individual unit of output, these things can enable a company to move closer and much quicker to the goal of maximizing ROI through automation. Process mining has always been a mainstay of process automation, but the surprising part – it is often overlooked or not given the priority it deserves by clients in their rush to RPA and Intelligent Automation. The underlying assumption is that processes in the company are running in the manner BPM has mapped them out, without exceptions. But in reality that is often not so, there are many alternatives and exceptional paths that business in a company gets executed, and it is the goal of process mining to find them all out and present them to the company’s process engineers for inspection and further corrective action. The good part, process mining is a leading business process management technique that can be used relatively easily and quickly by an organization without any disruption to the underlying business processes. And the even better part, it can be used by companies of all sizes, it is not exclusively a large company technology. Once implemented it allows for ease in process automation and provides much better results than any manual mining operation. The purpose of this short blog is to lay out the 5 most important factors that indicate your company really needs to look into process mining as the way to better understand your business processes. If your response to any one or two of these 5 key process characteristics is in the negative, our recommendation is your look into process mining. In our view, it is very likely your business process will benefit big time from process mining, most companies we are aware of do! It provides real-time, actual, quantitative guideposts of your processes – that is in itself is a big advantage as all companies are heading to complete digital transformation. Moreover, it will allow you to analyze which specific processes are most in need of automation, and if using process mining can improve the company’s processes substantially and the order with which to begin! Here are some critical signs to look for in a process that suggest your process likely will be in need of process mining, and the benefits it provides: 1. Significant Data Analysis in Your Process By combining Robotic Process Automation (RPA) with Process Mining, businesses are now able to configure massive amounts of data and highlight the most valuable data and trends in order to make it readable to employees. So, keep in mind, process mining not only helps to identify and eliminate inefficiencies but it also streamlines process data collection. If you are getting overloaded with process data streams, and you’re able to only look at a few, process mining can help a great deal in this regard. Below is a data sample from SAP Process Mining by Celonis. Not only does their process mining tool streamline the data but also enables the company to customize their own categories for data sorting. 2. Not Truly Understanding your Process Not having a full understanding of business processes, and all the ways business things actually get done, can a big problem especially in the face of imperatives such as 100% process automation and digital transformation. This lack of process knowledge can come from many different factors: in some cases, processes are just too complex to follow so employees generalize them and hope for the best. Over time, this can lead to discrepancies in how each worker understands the processes and how they actually work, since essentially no one is fully aware of the process as a whole, rather everyone only knows it from their specific view of the surface. If it is hard to follow what is actually being going on in your processes, and in particular you can’t get a hold of all the exceptional process paths that works get done on, that is a surefire sign that process mining is likely needed and will be of great help. Take a look at the Purchase-Order example below and you will see how crazy and complex processes can be. More often than not, business processes tend to be just as detailed as the one below, if not more. Without a process mining tool, businesses can never truly wrap their head around the ins and outs of their process, causing them to miss the small deviations that go on to cause greater issues later on. 3. Complex HR work The HR department tends to contain a plethora of repetitive tasks that can frustrate humans. All the manual labor that goes into inputting information into spreadsheets and logs is not only inefficient but is not making the best use of the people in HR whose jobs pertain to communication with others. Process mining (especially the element of it called task mining) can allow for these employees to be doing the work they were meant to do while their tedious processes become essentially autonomous. 4. Unidentifiable Problems or Gaps in Your End-to-end Process If businesses do not have a clear picture of their processes end-to-end as they actually exists, they will find it difficult even to notice or pinpoint any serious issues, let alone find solutions for them. Not understanding the process end-to-end can mean that there are gaps needing to be understood and filled in. Within these gaps are where the most bottlenecks are likely to arise since there is a lack of accountability for what the process flow is. Process Mining

Category: Robotic Process Automation

Is Your Company Ready For Process Automation?

How Machine Learning & AI Are Transforming The Future

The Microsoft Power Platform Tool Set: A ‘Down-to-Earth’ Primer

Automation of An External Data Feed Extraction and Assembly Process

Microsoft Power Automate Desktop: Capabilities and Instructive Scenarios on How to Use It

Analytics & RPA: The Transformative Digital Enabler

Incorporating NLP Capabilities into Innovatix’s WebMineR

Common Causes of Failure Among Process Mining Initiatives, and How to Avoid Them

Document Understanding: A Short Guide on the Major Process Steps

Is Your Business In Need of Process Mining?

Our Latest Post

Top 10 Custom Software Development Questions

Tips On Choosing The Right Staffing Agency

There’s No Such Thing As Too Much Testing

Categories

Our Latest Post

Categories

Tags