HGS heavily uses AI and ML, internally and in client facing solutions

The Coronavirus pandemic has tremendously increased the load on various customer service channels. As per my estimate, with our clients, between the call centre and the website, the call volume has shot up by over thirty percent. The calls received by our clients are relatively higher in the ‘critical’ category, as compared to pre-covid times. They have to be swiftly and efficiently responded to. It not only results in a successful call resolution but a happy customer. In order to meet the unique demands emanating from the challenges from coronavirus, Hinduja Global Solutions (HGS) has undertaken a number of initiatives both internally and in the client facing solutions, having a heavy dosage of AI and Customer Experience management (CXM).  

In the pre-covid times too, customer experience and management was a big focus area for HGS. The Coronavirus pandemic has resulted in a severe fast forward of the digital initiatives. Hyper-personalisation is the key theme coming up in the area of CXM. Ecommerce websites have to match the potential likes of the individual customers browsing products on the website and not bucket them in a customer segment. Another emerging theme is social media integration into the company website. The customers are reaching out to the companies through various social media channels. The queries have to be answered after having a look at the posts on various channels. HGS is heavily using AI and ML in the various offerings by the company.

AI in hiring
HGS is using the Amazon model of testing the solutions internally in the company before releasing it for the clients. For example, now employees are hired, onboarded and served over applications which are AI enabled. We are heavily invested in contact centres & back office operations. One of the major challenges facing the industry is attrition. It’s in the absence of a right hiring strategy that new hires leave in as few as under three to four months after joining. We have empowered hiring solutions with AI and CX features. The candidates are hired not only on the basis of an interview and technical tests but after exploring multidimensional datasets like demographics, past jobs, assessment scores & personality traits, etc. A right-fit index has been devised for candidates to evaluate whether they are the right fit for the job or not. This initiative has led to a 10-15 percentage point reduction in attrition at HGS.

After the employee has joined, the first three months are extremely important to determine the job satisfaction level of the employee. In order to make sure that an employee doesn’t have three bad calls (in that case he is asked to leave) HGS extensively uses AI to gauge certain early warning signals to indicate whether the new joinee is stressed, happy, requires training interventions or a different supervisor, or should be moved to a different project, etc.   

AI in Client facing solutions
The company is using AI and NLP in the insurance industry at a huge scale. The subject of area here is claims processing. On an average, ninety five percent of claims get approved and the rest get rejected. But if one looks at the rejected claims, many of them are asked for review. Every review claim is about 20-30 pages long. This amounts to thousands of pages on a regular basis. AI and NLP come to the rescue here. The technologies help in contextualising the sentences in the review claims, wherein every word is put into the right context according to what it means in the insurance industry to arrive upon a decision whether the review should be approved or if the claim still stands rejected.

AI is heavily used in quality checks of the call centre agents. Hitherto, only about two percent of the calls were checked for quality. The inability of doing a sentiment and sentence context check of the calls was a lost opportunity. Now, using AI, we are able to do a voice to text of all the calls and AI is run over the text to do not only quality check of all the calls but it also helps in arriving upon certain best practices on how to pitch to the customer.

Another service offered is Cognitive Contact Centre as a Service (CCCaS). Earlier an increase in the number of calls would result in additional hiring. No longer. We have created bots which can answer calls, which are easier in the kind of questions asked. In case if the bot is unable to resolve the call, then it goes to an agent. However a knowledge repository of such calls is created for the bots to learn. It prepares them to answer for similar calls in the future.   

Hinduja Global Solutions (HGS) is working with a retail giant in the US, on email automation. The company is flooded with customer emails and it’s impossible to manually reply to each mail. We use AI, ML, NLP to respond back. In a modular approach. Some emails are replied to in an automated fashion using AI, while the others are followed-up on a case by case basis. Emails are also segregated based on certain keywords and action is taken accordingly. In the last eight nine months (corona period), this customer engagement approach has hastened to unbelievable extent.

The vision of HGS is to compulsorily have a CX and AI component in all the solutions and also to increase the net promoter score (NPS).

Abhishek Raval also contributed to this story