7 Use cases in Intelligent Process Automation that Transformed Client Businesses

The world is currently in an unexpected conundrum due to the pandemic that has caught all organizations off-guard and crumbling economies globally. However, this situation has accelerated the adoption of Intelligent Process Automation.

Enterprises globally were made to reimagine their businesses and future of work, however, there’s one thing many of us did not see coming – rapid digital transformation that saw years of innovation under 6 months (and in some cases less). Infact we saw companies shifting and technologies getting adopted almost overnight.

As systematic time management of these crisis stuck enterprises was affected due to disparate working environment, poorly executed RPA, integration conflicts between legacy systems and new software applications and other such discrepancies lead to a non-optimal performance despite digitization for some. And the solution to this lies in complete business process transformation – the digital way of working which encompasses technologies such as Artificial Intelligence, Machine Learning and Natural Language Processing.

This is where Robotic Process Automation (RPA) – the effective technology to automate repetitive tasks becomes the saviour while its smarter sibling, Intelligent Process Automation (IPA) like the name denotes has cognitive capabilities that enables it to handle varied kinds of tasks using artificial intelligence.

In Laymen’s term, RPA is a rule-based and trigger-driven technology to automate repetitive, high-volume tasks. (Used for mimicking user steps without making any radical changes to the process). With the advent of technologies like NLP, ML, AI, and Cognitive OCR we now see the gear shifting towards IPA having the intelligence and ability to learn from previously programmed decisions.

The Paradigm shift from RPA to IPA (Intelligent Process Automation)

While many companies and enterprises are exploring innovative avenues, most of them have a piece-meal approach i.e. automating only one process or a few tasks at a time. There are complex challenges that hold them back from marrying together human and digital workforce (bots with brains) to run smarter operations.

As intelligent digital solutions and sustainability-driven technology choices help enterprises remain resilient, customer satisfaction is of utmost importance and so the strategy behind automation should be to elevate experience and not govern them one process or task at a time. At Hexaware, we have cannibalized our own revenues in multiple client engagements proactively, understanding the fact that embracing automation makes, not only us but our clients and their end customers, future-ready and in the long run we all stand to benefit mutually from the ROI.

Creating enterprise-wide automation model

For a Belgium based banking and insurance major, we implemented Enterprise-wide automation with a process bot factory. The firm was plagued by manual processes leading to dependencies and their resulting errors, demand fluctuations, and legacy apps slowing the processes further. Hexaware saw room for improvement, but quickly realized that traditional development methods weren’t working, and this called for modern, innovative strategy. Hexaware started with a simple document and workflow management approach to move on to create a crucial enterprise-wide automation model.

We also hold the record for having institutionalized an onsite-offshore scalable Automation Factory model with a Fixed capacity approach, enabling a high degree of predictability and quality for automation delivery by committing to 1 bot per month delivery. Additionally, we also implemented ‘Attended Automation solution with Human-in-the-Loop’ approach to address incoming queries and process transactions, with real-time Bot guided activities.

7 real-world use cases of Intelligent Process Automation (IPA) implementations

Listed below are successful automation implementations which produced proven results from an enterprise standpoint:

1. Automating warranty claims

For the world’s premier global providers of warranty solutions, the scope of engagement included an end-to-end automation of vehicle insurance and warranty claims management.

Unstructured inputs related to the damaged vehicle is received from the repairer for claims settlement led to an increased complexity for the RPA platform to handle. We proposed and implemented a cognitive automation solution that incorporates NLP, Image Analytics, Machine Learning and Robotic Process Automation tool (on UiPath platform) to automate the warranty claims process for the customer. Our AI/ML engine checks customer complaints, repairer comments, contract terms  and assesses the vehicle image to identify the damage type on the basis of which it takes the decision on whether to approve or reject the claim.

Our Automation solution explored the potential of NLP and deep Convolutional Neural Networks (CNNs) to transmute the process and leverages image classification and text classification while processing a claim to determine whether to accept or reject a claim.

This in-turn helped to improve the overall through-put when it came to Claims management. Almost ~95% of the volumes were processed by ML engine and areas where the machine had low confidence, AI can perform a contextual hand-off to a human agent. While agent processes the query, ML will learn from such scenarios to perform on its own if similar case occurs in future.

2. Adoption of automation in background verification

For a leading background verification company in North America, recordings of conversations between the agent and customer were saved and manually transferred to the QC folder after every call. This saved transcript is then picked up for quality check where excessive time was spent on performing QC on speech recordings to identify the quality of the call and flag any anomalies found.

With Hexaware’s Intelligent process automation solution, live calls are converted to text using Google Speech to Text API and stored for the RPA bot to pick up the transcript and notify the QC reviewer. The conversations from the source system (call recorder, VoIP stream) and the corresponding data such as the agent handling the interaction, date and time of the conversation and the customer details are included with each instance. Once the audio undergoes the speech recognition process where voice is converted into text, performing QC on text transcripts are faster than speech recordings. This solution, juxtaposed on our current DVaaS model, provides a unique approach when it comes to processing cases and in-turn boost customer experience not only for our customers but customer’s customer as well.

3. Contract management workflow automation

For a leading European Pharma Company, contracts received were severely unstructured in terms of their content; with financial and reputational risks involved in terms of non-adherence.

Hexaware’s Intelligent automation solution radically transformed the contract management process. A blend of Email Automation, Machine Learning based OCR and RPA solution has helped the customer create each contract in less than 30 minutes, reducing the AHT by 60%. Our custom-built Python engine leveraged Natural Language Processing (NLP) to understand the content of the contract and integrate it with the RPA solution to extract incoming contracts from emails and other sources for processing. Cognitive OCR tool has helped in defining multiple templates and layouts to capture data from multiple formats of structured and unstructured contracts and thus created an output of relevant details to be fed into their proprietary system for contract creations resulting in 100% closures the same day.

This meant, bring down the actual AHT which would take several days and transforming; the contract dominant organization; in totality.

4. Cash Application Process Automation

For a leading manufacturer of building products headquartered in North Carolina, business processes and workflows were majorly manual with the existing legacy system creating hindrance in scaling up while lacking speed and agility. Unstructured input templates in different formats across geographies also added to the complexity.

Hexaware proposed and implemented a Cognitive Automation solution that incorporates NLP, Expert Systems, Python scripting and the Blue Prism Robotic Process Automation tool to automate the Cash Application process for the customer. NLP based Entity Extraction was used to identify and extract the customer details provided in the bank statement narrative. It could identify details like customer names, customer reference numbers, invoice numbers and quote numbers from the unstructured text. The Expert System was built to parse the remittance files and extract invoice details. The Remittance Extraction Engine handled multiple formats: excel, pdf, doc and html and a wide range of remittance templates received from different customers.

The solution resulted in 42% saving on manual efforts through Intelligent Process Automation using NLP, ML – Python & RPA interventions

5. Automation in Trade Finance

One of world’s largest banks headquartered in US was experiencing gradual erosion of revenues in cases of discrepancy for the bank issuing Letter of Credit and there was a visible delay in processing leading to Financial/ Regulatory risks.

According to the report published by The International Chamber of Commerce (ICC) titled Global Trade– Securing Future Growth in 2018 trade flows are predicted to reach US$24 trillion by 2026. This meant that successful implementation of Intelligent Process Automation in Trade Finance would help organizations gain exponential cost benefit.

A solution was created for document checking of the unstructured Letter of Credit and discrepancy management process for negotiating the documents and providing post shipment facility. Data from unstructured paper documents and Letter of credit and Swift messages was extracted using RPA + OCR. Using NLP and Python, a Machine Learning web app was created for LC review and approval process involved further verification based on review of available supporting docs. The benefits delivered to the bank included 70% straight through processing along with considerable savings in terms of time, efforts and resources.

6. Automation in Inventory management

For a Global Beverage Giant, the current process required people to manually analyse 1000s of images per month from retail shelves and vending machines to analyse and place orders for restocking. This was not only manual intensive but time consuming too.

Hexaware proposed a solution where processing of the Image was done to automatically to extract the SKUs, track the sales and stock available for client in vending machines / super market racks. CCTV images combined computer vision and Python enable easy transference of images. It will then go to infer the images and provide analytics and insights on available items and whether it will need restocking based on the consumption pattern. This will also provide an insight on whether to order any item in surplus to meet the demand or discontinue the said items from the store. We also supplemented the solution with an option to be reordered automatically when the stocks are getting depleted without the need of store supervisor to manually check the stocks.

7. Automation enabled Front Office & Back Office Integration

The commercial cards space expects faster resolution for its card holders however the disparate processes and inter-departmental handoffs were pulling down the AHT considerable ultimately impacting the customer experience for one of world’s Largest Banks headquartered in US. Even for the largest of organizations having strongest front office and back office infrastructures will be siloed.

Hexaware started by gaining insights on the end customer composition, which was heavily dominant by boomers along with fast growing percentage of tech savvy millennials, It was pertinent to understand this volatile mix because the ask for each is different and hence should be serviced differently and this is where the flexibility of IPA comes to play. We integrated customer facing applications with traditional RPA coupled with Machine Learning and Natural Language Processing solutions to classify and extract the information to create a ticket. The bot then logs into client application and makes the requested change and closes the loop by sending a confirmation to the requestor. Its Human counterpart address complex request and help boomers navigate the tech landscape.

This automation enabled Front Office-Back Office Integration Model has helped bring down the turnaround time for email request by a whopping 90%. The outcome is twofold; Human element to cater customer who find it difficult to wade through the technical jargon and machine learning enabled solution will continuously learn from these exceptions to significantly bring down such cases in future especially for the rising percentage of millennials. The solution was equipped with Sentiment analysis to have a deeper insight on customer purchase pattern. This solution was built keeping in mind the shifting needs of the end customer and the growing number of digital platforms.

With Intelligent process automation (IPA) as a key enabler, your technology overhaul should cover three basic principles for enterprise-wide success and automation adoption:

• A multifaceted strategy keeping your customer as fulcrum
• A well-defined plan encapsulating your strategies with vision and organization level goals
• Bringing the plan to life to unleash maximum revenue potential and transform customer experience

A holistic approach will also allow you to keep an eye on revenue draining processes and metamorphose your organizations true potential.