Large organizations as part of their Digital Transformation are focusing on Business Process Automation in order to reduce costs, relieve employees of repetitive low value-added tasks and evolve towards an optimized and scalable process design.
According to Deloitte’s annual global executive survey, 73% of organizations worldwide have already embarked on the path to Intelligent Automation, but only 13% of them have succeeded in doing so at scale .
There are a multitude of technologies and technology service providers but each company must know them in depth in order to understand which ones are best suited to their core needs, and thus ensure that they can go beyond simple proof of concepts (POCs) and be able to execute a complete Digital Transformation.
Document Process Automation
In the world of Processes with high volumes of document classification and data extraction workloads, Automation is especially interesting, as it represents a great cost for companies in sectors such as Banking Services, Insurance, Legal or Public Administration. Such companies require large back-office teams that are inflexible to changes in demand and process requirements and bottlenecks are common, resulting in high variability of Average Operational Times per operation.
We can summarize in 3 main reasons to prioritize Document Process Automation through Artificial Intelligence;
Costs
Through the use of Natural Language Understanding (NPL) techniques, Computer Vision and training systems based on Artificial Intelligence, recurrent reductions of up to 50% in document classification and data extraction costs are achieved.
Response Times and Scalability
Under a flexible and scalable work system (Cloud or On-Premise) it is possible to drastically reduce the Average Operational Time per operation as well as to adapt to sudden volume growth.
3. Reduce Risks and Errors
And last but not least, the error is reduced to <1% which in many cases is a very considerable improvement over traditional manual focusing.
Limitations of Current Automation Approaches (RPA and OCR)
In general terms, one of the first initiatives promoted by large companies in the field of Automation has been the deployment of Robots in RPA (Robot Process Automation) systems. In short, RPA systems emulate human behavior, reducing the burden of simple, repetitive, low-value tasks. They are systems that offer good results in processes such as, for example;
– Copy and paste structured and machine-readable data.
– Click and drag files and attachments across multiple file paths.
– If/then decisions such as “if a PDF document is uploaded, then
Use it as input on platform X”.
– Opening e-mails and attachments without an understanding of the subject matter
of the context or content in the e-mail or attachment.
– Making calculations
When it comes to processing large quantities of documents, the evolution to the manual approach of BPO companies or internal back office teams has been OCR (Optical Character Recognition) systems. These systems transform images into text on a character-by-character basis using mainly basic rules and pre-trained templates. That is, it requires users to train the system for each new document type and only works with document types containing structured data.
In short, both systems (RPA and OCR), which usually work together, are not effective in terms of automation and hit rates when working with semi-structured or unstructured documents and data, complex workflows and high volumes.
According to Gartner, 80% of documents and data in large companies are unstructured, a figure that is increasing by 35% year on year. That is why large companies need an Intelligent Automation system, capable of automating complex processes with all types of documents and data, that is flexible and scalable and does not require continuous maintenance.
About Serimag
Serimag is dedicated to helping large companies implement new technologies that improve their processes and productivity.
The company focuses on the automation of business and back office processes through its TAAD Intelligent Automation system using the latest technologies in Artificial Intelligence, Computer Vision or Natural Language Processing.
Figures and customers
Serimag processes >1 Million pages per day for different customers in more than 80 production processes. TAAD recognizes >200 document types and extracts >400 fields, achieving an average project automation of more than 80%.
She has been specialized in critical processes in the banking and insurance sector for more than 10 years with very important clients at national level such as Caixabank, BBVA, Santander or Arag.
As a technology service provider, Serimag adapts to each project and client, integrating its TAAD system into the client’s systems and flows (ERP, RPA, BPM…) and deploys the project in the environment that best suits their needs (Cloud, On Premise or Approved DPC).
Human in the loop (BPO + TAAD)
Finally, in addition to providing the Process Automation part, Serimag offers a comprehensive service with its own BPO 2.0 In-House department, which performs a manual completion of the non-automated process through Serimag’s manual debugging applications.