Intelligent Automation

From automating mundane rule-based tasks to performing cognitive tasks

such as perception, reasoning, learning, etc., Intelligent Automation is disrupting traditional ways of improving operational efficiency, cutting costs and enhancing overall customer satisfaction across the value chain. The focus of Intelligent Automation solutions is to mitigate the ‘value leakage’ that organisations currently face in their front/middle/back office functions due to disparate systems, data complexity, highly manual and error prone tasks, non-standard processes, etc.; enhance the overall efficiency and effectiveness of the function. Therefore, leading organisations across industries have gauged the potential of Intelligent Automation as a key catalyst in enabling a paradigm shift towards ‘smarter’ organisations.

The spectrum of Intelligent Automation (IA) is classified into three classes:

Class 1 - Basic Automation

Uses Robotics Process Automation (RPA) tools which are typically designed to operate on the presentation layer of business applications without interfering with the underlying IT architecture. It processes structured data (spreadsheets, data present in relational databases, etc.), rule-based and transactional tasks by mimicking human actions.
RPA solutions are easy to implement and also observe faster benefit realisation. However, gauging the right candidates for RPA is imperative to yield maximum and early benefits.

Class 2 - Enhanced Automation

The financiers using calculator calculating personal taxes with colleagues in office workplace.

Uses data extraction techniques augmented with machine learning capability to ingest unstructured data (scanned document images, PDF and scanned handwritten images) with a higher accuracy rate and confidence level as compared to traditional OCR (Optical Character Recognition) tools; perform business rules/validation which are required to make sense of the data.
The machine learning component enables pattern recognition (context identification, cluster formation, label identification, etc.) through supervised/unsupervised learning of unstructured data. This results in a higher throughput rate of the desired output with increased efficiency and effectiveness.
Enhanced solutions are easy to implement. However, it typically requires 3-4 months of lead time for the engine to produce higher accuracy and throughput rates

Class 3 - Cognitive Automation

Uses sophisticated AI technologies such as Natural Language Generation, Speech Recognition, Computer Vision, etc., to ingest super data sets and perform cognitive tasks previously done by humans such as reasoning, perceiving, interacting with the environment variables and problem solving.
Cognitive solutions incorporate advanced self-learning capabilities, and can be used for sophisticated cognitive hypothesis generation/advanced predictive analytics.
Such platforms are costly to develop and implement, and generally require a long lead time. It reduces human error, but does not take humans out of the equation.

Business people hand pointing at business document during discussion at meeting.

How to get started with your Intelligent Automation journey?

Some of the biggest challenges that organisations typically face in adopting automation are uncoordinated efforts between stakeholders, inability to build a realistic business case, determining the right set of opportunities to automate, deploying the ‘right solution’ in the ‘right way’ and addressing change management issues post implementation.

KPMG in India will help you embark on the automation journey with its broad range of services in core consulting and implementation, with its deep expertise across the Intelligent Automation lifecycle.

Below are the broad service areas and its offerings:

Core Consulting

List of offerings under Core Consulting

Implementation and Support

List of offerings under Implementation and Support

IT Architecture and Security