Robotic process automation is an effective way to automate repetitive tasks that involve structured data, such as from spreadsheets and databases. The problem is about 80% of data is unstructured, including PDFs, Word documents, emails, images, videos and more. Intelligent process automation has cognitive capabilities that enable it to effectively deal with unstructured data, Combine IPA with RPA and now you have a way to automate tasks and workflows that involve both structured and unstructured content. In short, RPA and IPA are complementary technologies, not competitive.

Download the Everest Group Whitepaper: Data Process Automation

Download Now

Robotic Process Automation (RPA)

  • Automates deterministic, repetitive processes involving structured data.

Intelligent Process Automation (IPA)

  • Has cognitive capabilities and can automate processes involving unstructured data, including images, PDFs, emails, Word documents, and more.

RPA + IPA

  • IPA ingests unstructured data, converts it to a structured format, feeds it back to an RPA tool.

Result: Businesses can automate processes involving both structured and unstructured data, realizing:

image shows that Indico's intelligent process automation can lead to an 85% reduction in process cycle time

85% Reduction

Process Cycle Time

image shows that indico's intelligent process automation can lead to a 4 times increase in process capacity

4x Increase

Process Capacity

image shows that indico's intelligent process automation can lead to an 80% reduction in the amount of total resources required

80% Reduction

Resources Required

Robotic Process Automation Explained

RPA involves automating repetitive tasks to make a process less labor-intensive for humans. It works well with deterministic business processes that involve structured data; in other words, where the process is exactly the same every time and where data is in well-defined fields, such as a spreadsheet. The process also must not require any human judgment, working strictly on “if/then” scenarios.

With RPA, users program a software robot to follow the steps a human would normally take in performing a process. The robot performs tasks that are repetitive – and boring – for humans. By using computers to perform the tasks instead, RPA lowers costs and promises better accuracy, because computers don’t get tired or make keystroke errors.

RPA involves automating repetitive tasks to make a process less labor-intensive for humans. It works well with deterministic business processes that involve structured data; in other words, where the process is exactly the same every time and where data is in well-defined fields, such as a spreadsheet. The process also must not require any human judgment, working strictly on “if/then” scenarios.

With RPA, users program a software robot to follow the steps a human would normally take in performing a process. The robot performs tasks that are repetitive – and boring – for humans. By using computers to perform the tasks instead, RPA lowers costs and promises better accuracy, because computers don’t get tired or make keystroke errors.

Common RPA Use Cases

Sales functions

Customer relationship management systems are useful tools to help salespeople stay on top of customers. But it’s also time-consuming for sales and marketing professionals to keep them up to date. RPA can help, by automating the process of extracting customer data from invoices, purchase orders and other systems and entering it into the CRM system. Here again, so long as the fields the data is coming from and going to are well-defined, RPA will be up to the task.

Tech support

In a tech support scenario, RPA bots act as a first line of contact. They can help solve simple issues, like password resets, as well as diagnose issues by asking a series of questions. When issues do need to be escalated, the human support agent will have some preliminary information and be able to get right to the job of diagnosing the problem and helping the user.

Financial reports

Example use cases for RPA include aggregating data for financial reports, such as at the end of a quarter. So long as you know what reports the data is coming from, and where in each report it’s located, RPA can automate the gathering and aggregation process and get it done far faster than a human. In banking, RPA can automate the process of copying and pasting customer data from one banking system to the next. For credit analysis, RPA could automate the process of logging in to a credit bureau portal, uploading customer details, and downloading resulting credit reports.

Intelligent Process Automation Explained

IPA solutions can automate processes that involve unstructured content, which often accounts for about 80% of all content enterprises deal with. IPA builds on the AI concept of transfer learning, which enables a model trained on one task to be used to perform another, related task.

As this white paper from the Everest Group explains, IPA solutions also support technologies including:

  • Machine learning (ML) and deep learning models to classify and extract documents, and perform software training.
  • Optical character recognition (OCR)/Intelligent Character Recognition (ICR), to convert document images into machine-coded text, using ML and deep learning algorithms to train for increased accuracy.
  • Natural Language Processing (NLP), to analyze text in documents, understand surrounding context, consolidate extracted data, and map the extracted fields to a defined taxonomy.

These capabilities enable IPA solutions to learn over time and give them cognitive capabilities to handle some human-like decision-making that can be applied to all sorts of document-based processes. For example, an IPA platform could examine a set of RFPs and score them according to how well they meet business objectives. For insurance claims analysis, IPA solutions can quickly examine hundreds of claims and identify those that may indicate fraud. In summary, IPA does not replace or compete with RPA, it complements it. IPA translates the unstructured content into structured data so it can be plugged back into the RPA platform. And IPA is already being applied to a number of common back-office use cases related to legal & compliance, sales & support, and finance & operations, and more. Following are a few examples.

IPA solutions can automate processes that involve unstructured content, which often accounts for about 80% of all content enterprises deal with. IPA builds on the AI concept of transfer learning, which enables a model trained on one task to be used to perform another, related task.

As this white paper from the Everest Group explains, IPA solutions also support technologies including:

  • Machine learning (ML) and deep learning models to classify and extract documents, and perform software training.
  • Optical character recognition (OCR)/Intelligent Character Recognition (ICR), to convert document images into machine-coded text, using ML and deep learning algorithms to train for increased accuracy.
  • Natural Language Processing (NLP), to analyze text in documents, understand surrounding context, consolidate extracted data, and map the extracted fields to a defined taxonomy.

How IPA Complements RPA

Corporate inbox

Most companies have a central inbox that receives lots of email from customers, contractors, suppliers and the like, often with attachments. RPA can be used to detect when a new email arrives with an attachment, then automatically route the email to an intelligent automation tool. The IPA tool can then extract the attachment and “read” it, using OCR and NLP. It can extract relevant unstructured content such as payment terms, invoice numbers, contractual language and so on. The tool can then normalize the data in an appropriate format and send it to a downstream platform, such as a customer relationship management (CRM) or enterprise resource planning (ERP) tool.

Contract renewals

RPA platforms can automate processes triggered by a specific event, such as the end of a contract period. A cable television company, for example, could use RPA to automatically send customers an email when their contract is nearing its expiration date, urging a renewal. But with its ability to understand context, an IPA tool could look at the customer’s current service lineup and activity throughout the year, such as movie rentals. In the process, it may determine whether there’s an opportunity to upsell additional services or offer attractive packages and add appropriate language to the email.

Invoice automation

For invoice processing, RPA can automate data input, reconciliation error correction and some decision-making. But the challenge is dealing with the many formats different vendors use for their invoices. That’s where the IPA platform can contribute, by creating an extraction model to pull out necessary data from the invoices, normalize it to a structured format, and send it back to the RPA platform for automated data input, error handling and so on.

Financial document analysis

Financial firms need to compile lots of data for monthly and quarterly reports. RPA can aid in the process by automating data collection from various structured sources. But introduce an unstructured PDF document to the process and RPA hits its limit; now you need the OCR and potentially NLP capabilities of an IPA solution to pull out relevant information and convert it to a structured format that the RPA tool can deal with. (Learn more through this financial services IPA use case).

Insurance claims

Insurance companies can automate some aspects of the claims process with RPA platforms, such as inputting data from structured sources and ensuring all required fields are filled out. But insurance claims often include unstructured data, including photos showing auto damage, PDFs of scanned driver’s licenses, or perhaps images such as CT scans for a healthcare insurance claim. An IPA platform can be used to extract relevant information from these sources, once again adding value to the RPA tool.

RPA + IPA Delivers Big Benefits

Complementing RPA projects with IPA technology means you can now get all the benefits that IPA delivers, including:

85% reduction in process cycle times

Drive customer satisfaction and quicker time to market for new initiatives

4x increase in process capacity

Scale critical processes without increasing expenses, for more cost-efficient back office functionality

80% reduction in human resources

Free up critical resources to work on higher value-add projects rather than repetitive low-value tasks

1000x less training data required

As compared to traditional artificial intelligence solutions