2 minute read
In fast-paced corporate environments, AI, powered by linguistics and computer science, is taking over all the tasks that can be automated. The trend is expanding fast to businesses in all areas, from hospital and HR companies to private schools and universities. Natural language processing undergoes continuous improvement to the extent that communication with a chatbot is hard to be distinguished from exclusively human interaction. At the NAACL Conference in 2019, several financial institutions and organizations shared their experience in implementing AI in their customer-facing routine. Umpqua Bank integrated AI in their hiring and onboarding process. From application to on-the-job support, there is little to no human interaction as far as the technical parts are concerned. Eve Callahan, EVP, Chief Communications Officer, underlined the enhanced experience both managers and new employees get. They used to struggle with the hiring process before setting up the virtual recruitment tools. They handle everything from screening resumes to scheduling an appointment at a time when the hiring manager is available. Eve Callahan shared that the mobile-first optimization is complete, and their next step is applying NLP to building more efficient customer pipelines. AI is already processing membership and loan applications at Ohio University Credit Union. The speech recognition software assists the visually-impaired customers in inquiring about their account details. The FAQ section has been restructured and adapted for the chatbot. Students of the Ohio University have drawn inspiration from the NAACL Conference tutorials and implemented an experimental lexical adaptation to the virtual loan assistant. The bot adapts to the applicant’s vocabulary and level of financial expertise when replying to questions. Every member can now easily understand details on loan terms, interest rate, APR and ask for additional details. If statistics show positive feedback, the project could revolutionize the work process of credit unions. BadCreditCashASAP handles a large number of loan applications. Their automatic scanning process ensures that no corrupted files go through, the sender is notified and provided guidance to correctly submit their documents to get a loan. On the other side of the screen, lenders only receive files that are ready to process. The bot assistant seamlessly integrated with the firewall designed to protect the personal data of all applicants and lenders alike. BadCreditCashASAP CEO Gregory Allen stated that the AI addition and system tweaks had improved their clients’ experience and the rate of successful applications raised by 19% in the first month, recording constant growth.