Automation lagging by over a decade in India: Sapient Consulting’s Raja Raman
Information technology firm Sapient Consulting has been looking to ease digital transformation of legacy systems globally by dividing its focus into two – Sapient Global Markets for banking, financial services and insurance, and SapientRazorfish for healthcare, transportation, e-commerce and others.
In an interview with TechCircle, Raja Raman, vice-president of technology at Sapient Consulting, says that India lags behind other countries in robotic process automation. Raman, who has been with the company for more than 17 years, also feels that such automation will become central to business operations.
Why is it beneficial for an enterprise to automate certain processes?
In order for us to understand the benefits of automated processes, we need to understand what robotic process automation (RPA) stands for or what it does. Think of RPA as a digital assistant behaving exactly the same way as you do, taking away the mechanical aspects of most jobs. In simple terms, it is a bunch of codes running together, with artificial intelligence (AI) and machine learning, which handles high-volume repeatable tasks that previously required humans.
The technology landscape has been changing rapidly over the last 17 years and has been expanding to more industries as cost of technology is coming down and businesses are seeing newer use cases. When I started my career, things were different – high cost of tech development, expensive hires and unavailability of ready talent were key issues.
Later, companies started using tech as one of the many facets of business operations or an extension to their businesses. But today, information technology is at the core of a company’s survival strategy.
If one doesn’t change over to the technology, there are enough digital disruptors in the market to take them on. Not only that, RPA provides the opportunity for companies to achieve significant cost savings and business efficiencies.
Let us take the example of a bank. Say Bank A provides 6% interest on deposits and 9% interest on loans. The 3% spread is what it uses for running its operations. Now let’s take a new-age digital Bank B, which runs primarily on digital properties; then it can offer 6% on deposits and 7% on loans, making it much more tempting for consumers.
What would be the typical journey for a company looking to adopt RPA?
A typical RPA journey can be observed in three steps: initialisation, industrialisation and institutionalisation. The initialisation phase mainly consists of continuous improvement and involves understanding of all components of RPA implementation. An initialisation phase may involve deployment of 10 to 15 bots across different functions, data-intensive processes and those involving simple validations.
The industrialisation step comprises of large-scale deployment of RPA and can include the automation of several hundred thousand transactions per month. Shared services, back-office operations and front-office support functions are good candidates for RPA in this phase.
Institutionalisation extends automation to all units of the organisation and automates the handoff between different business units (a handoff refers to the transfer of a data session from one channel to another. A well-implemented handoff is important for delivering uninterrupted service).
Successfully scaling up RPA depends on fully understanding and implementing all six elements of the robotic operating model. These six elements are: the potential percentage automation, the efficiency gains across different business functions, the different modes of interaction and touch points with other business functions and/or other business units, the human and bot interactions across different business functions and use cases, the different user journeys and user profiles in the current state, and target state across business areas and functions.
Has Sapient been able to show return on investment for a client after RPA implementation?
For one of our energy clients, we automated the processing of movement tickets received from different vendors with different methods of transport like trucks and railways. The vendors send the loading bill in different formats like email, PDF, scanned images, word document, and fax.
The bot uses the power of intelligence process automation to process unstructured and structured data sent in different formats, maps it to the system data and punches it into the downstream system of records.
Automated processing of bills has reduced processing time by 70% and overall the data accuracy has increased to 90%, thus reducing the overall reconciliation and correction work. Trading desk gets market intelligence data from varied sources across the globe.
This information needs to be processed and ingested into market intelligence system on a real-time basis. A bot monitors all the market intelligence data sources and as the information is received it does data filtration, interprets and enriches the data before punching it into the system. Automation of the process has helped the client in tripling market coverage and reducing errors by 95%.
How many types of RPAs are there or how can we classify RPAs?
We can mainly classify RPAs into three major categories. First, the basic one is rule-based which tends to solve a certain targeted problem and sits at the bottom of the value chain of investment. The second is the enhanced or digital RPA, which can solve more complex issues or tackle multiple functions. The third and the most-exciting one is cognitive RPA. These kinds of automation processes have a lot of AI built in. We are focusing on cognitive RPAs in Sapient.
A lot of rule-based kind of automation existed before the world moved to RPAs. In software testing, for example, there existed a lot of automation tools such as QTP (QuickTest Professional) and Selenium, which would help test applications. In this kind of scenario, to conduct tests you open a screen, type a bunch of commands on a web browser and click forms. QTP or Selenium monitors the entire process and can play it back to you.
But what happens in the back-end for all this to be possible is a whole bunch of codes working together. This led to the thought that if I can automate software testing, why can’t I automate business processes?
What is the global opportunity for RPAs and where does India figure on the list?
In terms of market spend on RPAs over the next five years, I see the world reaching low billions of dollars. India should be around half a billion but that also looks a stretch. The country simply has not reached a level of basic RPA adoption. Though there are some companies trying some solutions internally, we are just way behind. We are in the QTP (Quick Test Professional, a user interface testing tool) area, which would put us in the early 2000s.
I think the main issue seems to be that labour is quite cheap in India compared to the developed markets. But having said that, I see that India will start seeing a little more adoption in the basic rule-based space. Here, there are several jobs such as customer support, content aggregation and recruitment that can be automated to at least 60%.
So which market is Sapient betting on?
There is [a big] enough market for us to work on. The developed markets are yet to be fully served. Having said that, the US continues to be one of our key markets, with sizeable 10% of the revenue coming from it. As I said before, we intend to innovate on cognitive RPAs to lead industry-first solutions.