When rule-based automation arrived a few years ago, it came with a huge amount of hype. Robotic process automation (RPA)and its machine learning (ML)methods were going to revolutionise the way workers accomplished repetitive tasks, freeing them up for more high-value work and quickly saving their employers lots of money.
The reality hasn’t been amazing so far. In many cases, RPA was able to only help with the easiest parts of a worker’s job, so that it was the low-hanging fruit of highly structured data that was dealt with. Complex or completely unstructured data essentially remained a closed book; at that point the automation had no way to figure out what to do with it without lots of human assistance. It was for this and similar reasons that a 2017 study of automation buyers found that 41% of non-adoptors of RPA weren’t buying it because of a “lack of clarity on the business case.” Nearly one in four (24%) simply didn’t see the value to their organization.
Traditional rule-based automation often couldn’t do much without at least some analog help–it still needed assistance when dealing with the many variants and exceptions found in data, since such differences were hard to capture in the rules that governed a bot’s decision-making. This made RPA very difficult to scale, and instead of being freed up to do other tasks, workers had to frequently take on another duty: teaching the bots to deal with variations in the data. For many processes, this made it difficult to deliver savings, or even any improvement in accuracy or speed.
A blizzard of (unstructured) paper
The problem hasn’t gone away. According to the Future Ready Lawyer’s Survey of 700 lawyers, nearly three in four (72%) struggle with the increasing volume and complexity of document processing and want to focus on efficiency and productivity. It’s likely other knowledge workers have troubles at similar rates.
Every business of any size has to rely on unstructured (and in fact often non-digital) information. As highly regulated entities, banks and insurance companies have case law and policies and contracts to consult. Companies concerned with property rights, such as the oil and gas industry, hold extremely detailed agreements detailing the lands on which they hold rights and breaking down just what those rights entail. Similar documents cover the transport and maritime industries.
Such agreements and other documentation may go back decades or more, but still need to be consulted frequently, often by workers in the field who are far away from a desktop computer, let alone the documents themselves. Guidance from environmental authorities and other governmental agencies add yet another wrinkle to companies’ day-to-day functioning, as do the millions of emails, texts, PDFs, SharePoints, and other random documents that every company has to contend with.
To tame all this paper, which continues to be expanded every day, intelligent automation is really the only way out, because it scales up. Without it, more labor will continue to be needed to deal with it, organise it, and search through it, only making the problem worse.
Enhancing intelligence in AI
Today, semantic driven solutions have emerged to automate retrieving information from the unstructured chaos. Natural language processing (NLP)methods can extract the important intent and other salient points from contracts and other documents, and it does it without lots of (non-scalable) hand-holding, and with a high degree of accuracy, after being trained with a few documents.
“It is the combination of NLP and machine learning that will enable organisations to gain insights from unstructured data such as emails, chat transcripts, outbound marketing materials, internal memos, legal documents, and complaint logs, in a way that hasn’t been previously possible.”
– Jessica Davis, from InformationWeek (June 2019)
The benefits of NLP and other semantic technologies extend beyond just being able to ingest unstructured data: it can also allow workers to find information quickly, across an entire organization. And by using such skills, chatbots become much more effective, with ever-increasing accuracy in the answers and guidance they provide. To take one example of NLP’s gifts that you probably use every day, Google’s search results are now powered by algorithms that allow it to understand longer queries as well as ones in which prepositions and other “transformers” alter its meaning. This improvement, which Fortune magazine called “a big leap forward,” helps weed out useless or inaccurate search results. Similar advances are helping companies improve around the world.
It’s true that RPA’s initial promises tended to collapse under their own weight. But by taking a more sophisticated approach that exploits advances in NLP and semantics, companies are now finding ways to automate complex work that was impossible just a few years ago.
For more information about TAIGER AI solutions, click the button below to schedule a product demo with us.
AI & RPA – What’s the difference in reality and what’s next?
When rule-based automation arrived a few years ago, it came with a huge amount of hype. Robotic process automation (RPA) and its machine learning (ML) methods were going to revolutionise the way workers accomplished repetitive tasks, freeing them up for more high-value work and quickly saving their employers lots of money.
The reality hasn’t been amazing so far. In many cases, RPA was able to only help with the easiest parts of a worker’s job, so that it was the low-hanging fruit of highly structured data that was dealt with. Complex or completely unstructured data essentially remained a closed book; at that point the automation had no way to figure out what to do with it without lots of human assistance. It was for this and similar reasons that a 2017 study of automation buyers found that 41% of non-adoptors of RPA weren’t buying it because of a “lack of clarity on the business case.” Nearly one in four (24%) simply didn’t see the value to their organization.
Traditional rule-based automation often couldn’t do much without at least some analog help–it still needed assistance when dealing with the many variants and exceptions found in data, since such differences were hard to capture in the rules that governed a bot’s decision-making. This made RPA very difficult to scale, and instead of being freed up to do other tasks, workers had to frequently take on another duty: teaching the bots to deal with variations in the data. For many processes, this made it difficult to deliver savings, or even any improvement in accuracy or speed.
A blizzard of (unstructured) paper
The problem hasn’t gone away. According to the Future Ready Lawyer’s Survey of 700 lawyers, nearly three in four (72%) struggle with the increasing volume and complexity of document processing and want to focus on efficiency and productivity. It’s likely other knowledge workers have troubles at similar rates.
Every business of any size has to rely on unstructured (and in fact often non-digital) information. As highly regulated entities, banks and insurance companies have case law and policies and contracts to consult. Companies concerned with property rights, such as the oil and gas industry, hold extremely detailed agreements detailing the lands on which they hold rights and breaking down just what those rights entail. Similar documents cover the transport and maritime industries.
Such agreements and other documentation may go back decades or more, but still need to be consulted frequently, often by workers in the field who are far away from a desktop computer, let alone the documents themselves. Guidance from environmental authorities and other governmental agencies add yet another wrinkle to companies’ day-to-day functioning, as do the millions of emails, texts, PDFs, SharePoints, and other random documents that every company has to contend with.
To tame all this paper, which continues to be expanded every day, intelligent automation is really the only way out, because it scales up. Without it, more labor will continue to be needed to deal with it, organise it, and search through it, only making the problem worse.
Enhancing intelligence in AI
Today, semantic driven solutions have emerged to automate retrieving information from the unstructured chaos. Natural language processing (NLP) methods can extract the important intent and other salient points from contracts and other documents, and it does it without lots of (non-scalable) hand-holding, and with a high degree of accuracy, after being trained with a few documents.
The benefits of NLP and other semantic technologies extend beyond just being able to ingest unstructured data: it can also allow workers to find information quickly, across an entire organization. And by using such skills, chatbots become much more effective, with ever-increasing accuracy in the answers and guidance they provide. To take one example of NLP’s gifts that you probably use every day, Google’s search results are now powered by algorithms that allow it to understand longer queries as well as ones in which prepositions and other “transformers” alter its meaning. This improvement, which Fortune magazine called “a big leap forward,” helps weed out useless or inaccurate search results. Similar advances are helping companies improve around the world.
It’s true that RPA’s initial promises tended to collapse under their own weight. But by taking a more sophisticated approach that exploits advances in NLP and semantics, companies are now finding ways to automate complex work that was impossible just a few years ago.
For more information about TAIGER AI solutions, click the button below to schedule a product demo with us.
Search
Archives
Taiger’s CEO recently shared his insights on the current and future developments of Generative AI at a fireside chat hosted by ICEX
May 9, 2023TAIGER’s Omnitive IDP Solution Now Capable of Extracting Information from Vietnamese Documents
April 26, 2023Categories
Meta
Calender