Nedbank Insurance, part of the Nedbank Group that employs over 31,000 people across South Africa, has utilised Machine Learning and Design Thinking to help speed up how it processes emails.<\/p>\n\n\n\n
Challenge<\/strong><\/p>\n\n\n\n Today,\nevery business has an email address. And if you\u2019re a financial services\norganisation with many business sections and over a million clients, that\nchannel can get crowded.<\/p>\n\n\n\n Nedbank\nInsurance covers clients\u2019 life and other insurance needs and receives thousands\nof emails about claims, policies, address changes, complaints, queries and\npotential new clients. Some emails are long; others are short. Many have\nattachments that must be categorised correctly to simplify processing for\nback-end teams. There are also unpredictable peaks when for example heavy\nstorms cause inboxes to be bombarded.<\/p>\n\n\n\n Indranil\nBandyopadhyay, Head of Business IT Enablement for Nedbank Insurance, suspected\nthat an automated system could solve the problem and increase operational\nefficiency and client satisfaction.<\/p>\n\n\n\n \u201cWith\nthe old system, a single person can, on average, process one email every 60 to\n300 seconds (three minutes on average),\u201d he said.<\/p>\n\n\n\n \u201cThat\u2019s about 160 emails per dedicated resource a day. With email volumes growing, our backlogs were mounting, and we didn\u2019t want this to influence our client service. We knew technology could help us create a more sustainable solution, but we needed the right partner.<\/p>\n\n\n\n \u201cNedbank\nInsurance had to try a couple of times to get this right. The challenge was that they [other companies]\ndidn’t try to solve the real problem and looked towards their overseas\ncounterparts to solve for this problem. Some of them went half-way and gave up.<\/p>\n\n\n\n \u201cI\nalso engaged with international vendors who mentioned they couldn’t solve the\nend to end problem. This made me feel we are really trying this for the first\ntime in the world. I did not give up because I believed this could be solved.<\/p>\n\n\n\n \u201cGrit\nand tenacity along with the right partner were the reasons we were successful\nwith Synthesis.\u201d<\/p>\n\n\n\n Solution<\/strong><\/p>\n\n\n\n Following\nvarious methods to analyse email and attachment data, Synthesis built\npredictive Machine Learning (ML) models with natural language processing\nalgorithms to understand the intent of the email. The models first learnt from\nthe employees at a rapid pace, refining the process through feedback. Then,\nwhen the models performed reliably in terms of predicting where emails should\ngo, they were routed automatically.<\/p>\n\n\n\n \u201cI\nhave been thinking about ML for the past three years,\u201d said Bandyopadhyay.<\/p>\n\n\n\n \u201cThe\nbenefits are huge. Especially in an admin orientated industry like Insurance. <\/p>\n\n\n\n \u201cWhile\nRPA can deal with repetitive stuff, ML can be used where cognitive ability is\nneeded to complete a task. The main challenge of ML and automation in general\nis the introduction of ‘positive feedback loop’. This has a leg for a process\nto go rogue and can have huge negative impact in a very short period of time.<\/p>\n\n\n\n \u201cSometimes,\nthe amount of data to teach the ML algorithms is also huge challenge and the\nquality of data is also a major constraint.<\/p>\n\n\n\n \u201cBut\nwhat struck me is that Synthesis used Design Thinking to truly understand what\nthe problem was. In a very short time they created a prototype, tested it with\nusers and then adapted it. The solution they created revolved around genuine\nclient satisfaction and not just completing a task, and that worked for us.<\/p>\n\n\n\n \u201cDesign\nThinking assists in understanding the real problem and not the superficial\nissue. After asking 10 whys, one gets to the core of the problem. <\/p>\n\n\n\n \u201cThe\nissue for us was to find a solution to deal with a huge number of incoming\nemails. Our emails were in the Microsoft Inbox and the attachments needed\ndownloaded on to a local PC. <\/p>\n\n\n\n \u201cOnce\na human being understood what the emails were all about, they needed to go to\nour policy admin system and then categorised with the documents attached the\ndocuments.<\/p>\n\n\n\n \u201cWhile\nwe thought about using a ML\/AI solution, Synthesis was quick to understand the\nreal problem via DT. They quickly created a front end that pulled emails and\nattachments and, once the human being understood the content of the email, the\nfront end automatically attached the document in the right category. <\/p>\n\n\n\n \u201cThis\nimmediately gave 80% efficiency though no ML was used at this stage. This was a\nsimple yet powerful solution that the users started using and got immediate\nbenefit. This assisted in change management as well. At a later stage, we used\nan AI model to make sense of the emails and attachments and took the human\nelement out of the value chain.\u201d<\/p>\n\n\n\n Results<\/strong><\/p>\n\n\n\n The\nnew system has significantly reduced email processing time, on average from 720\nseconds to 28 seconds. <\/p>\n\n\n\n \u201cClients\u2019\nexperience will change immensely because now we can work on whatever our\nclients request immediately,\u201d said Bandyopadhyay.<\/p>\n\n\n\n \u201cDuring\na peak, the machine can handle the volume of requests and now nothing hinders\nus from responding to client needs. This will greatly improve client experience\nand will support our commitment to serving clients.<\/p>\n\n\n\n \u201cI\nam not surprised by how successful the solution has been because I slept, ate\nand drank this dream and was confident that this problem could be cracked.\nHowever, I was surprised to see the success rate from day one we went to\nproduction.<\/p>\n\n\n\n \u201cInternally,\nthere has been a very positive response. This solution is now showcased within\nthe group and there are many interested parties that would like to use such a\nsolution in this business.<\/p>\n\n\n\n In\nturn, Synthesis thinks Nedbank Insurance deserves recognition for its\npersistence. <\/p>\n\n\n\n \u201cThis\nis what it takes to pioneer advancements,\u201d said Marais Neethling, Synthesis AI\nEvangelist.<\/p>\n\n\n\n \u201cIt\nwill be great to support Nedbank Insurance as it continues its transformation\njourney and client satisfaction drive.<\/p>\n\n\n\n \u2018We\nhave significant experience in using emerging technologies to help\norganisations transform, but we\u2019ve found there\u2019s something more important than\ntechnology – applying Design Thinking. <\/p>\n\n\n\n \u201cWe\ntry to define an end-to-end system that is practical and includes all\nsupporting processes. <\/p>\n\n\n\n \u201cThe\nchallenge was interesting, and we had to work closely with stakeholders and\ncreate a plan with the Nedbank tech team.\u201d<\/p>\n\n\n\n Nedbank\nInsurance now wants to leverage the solution across the business. <\/p>\n\n\n\n \u201cAny\nelectronic interaction is the next step,\u201d said Bandyopadhyay. <\/p>\n\n\n\n \u2018All\nwe need to do when someone calls in is to convert the voice into text \u2013 and\nroute the call correctly. Artificial Intelligence and Machine Learning are the\nideal ways for us to tap into the digital world truly to meet our clients\u2019\nneeds. We are excited to delve further into the possibilities of this\ntechnology.\u201d<\/p>\n","protected":false},"excerpt":{"rendered":" Nedbank Insurance, part of the Nedbank Group that employs over 31,000 people across South Africa, has utilised Machine Learning and Design Thinking to help speed up how it processes emails. Challenge Today, every business has an email address. And if you\u2019re a financial services organisation with many business sections and over a million clients, that […]<\/p>\n","protected":false},"author":25,"featured_media":32133,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[565,34,338,843,328,19,183,9059,13,262],"tags":[496,386,12684,12685,392,12687,12686,12683,7531,10616],"class_list":["post-32130","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-banking-finance","category-case-studies","category-case-study-newsletter","category-editors-choice","category-features","category-regional-news","category-software","category-south-africa","category-top-stories","category-used","tag-ai","tag-artificial-intelligence","tag-design-thinking","tag-indranil-bandyopadhyay","tag-machine-learning","tag-marais-neethling","tag-microsoft-inbox","tag-nedbank-insurance","tag-rpa","tag-synthesis"],"acf":[],"publishpress_future_workflow_manual_trigger":{"enabledWorkflows":[]},"_links":{"self":[{"href":"https:\/\/www.intelligentcio.com\/africa\/wp-json\/wp\/v2\/posts\/32130","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.intelligentcio.com\/africa\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.intelligentcio.com\/africa\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.intelligentcio.com\/africa\/wp-json\/wp\/v2\/users\/25"}],"replies":[{"embeddable":true,"href":"https:\/\/www.intelligentcio.com\/africa\/wp-json\/wp\/v2\/comments?post=32130"}],"version-history":[{"count":4,"href":"https:\/\/www.intelligentcio.com\/africa\/wp-json\/wp\/v2\/posts\/32130\/revisions"}],"predecessor-version":[{"id":32137,"href":"https:\/\/www.intelligentcio.com\/africa\/wp-json\/wp\/v2\/posts\/32130\/revisions\/32137"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.intelligentcio.com\/africa\/wp-json\/wp\/v2\/media\/32133"}],"wp:attachment":[{"href":"https:\/\/www.intelligentcio.com\/africa\/wp-json\/wp\/v2\/media?parent=32130"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.intelligentcio.com\/africa\/wp-json\/wp\/v2\/categories?post=32130"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.intelligentcio.com\/africa\/wp-json\/wp\/v2\/tags?post=32130"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}