Like all businesses we are looking at the ways AI is going to change lending, our processes and what it means for our clients. Right now we are looking to train an AI for our chat window on this website. This process is showing me both the capabilities of AI and some of the draw backs. I believe it will be capable for answering general questions and interacting with our lending software in real time. But if you don’t know the right way to ask a question you may get an answer that is right but not 100% right (or fully considered). With this is mind we will always assign a human broker to work with you. If your needs were all about just numbers and transactional then I believe AI will work amazingly well. But 20+ years of asking clients about what they are trying to achieve makes me think that the technology might not be ready. When I ask a question, your response might make me ask 2,3,4 follow up questions or propose possible ideas/outcomes. That we then work through together so you can decide which is the right outcome for you. Oftentimes, this process leads to different outcomes from my initial thoughts on your requirements. This is the area that human interaction is hard to replace. As trust is built and a relationship formed. Your risk profile, future needs/goals all come into play. We want to provide you with a person, mobile number and direct email address should you have questions, doubts or additional ideas that spring to mind. Also not being reliant on an internet connection or server availability is still a strength in my eyes. While most infrastructures have very high “uptime”, they still do have downtime. Think Optus losing their network and eftpos being down around Australia. This was a significant problem for consumers and businesses alike, affecting millions of transactions from buying fuel, bread, milk to larger purchases.
We are starting to use AI to help us navigate the everchanging lender policies, rates & procedures. However, this can’t solely be relied upon. We still speak directly with bank staff to clarify policy and interpretation. No doubt AI will improve some of these experiences and potentially speed up response times and quality of information, but only if the right question is being asked. I still like to work through files with bank staff. As lending is not always as straight forward as people think, especially for people that are self employed or have complex income/lending structures.
I can see online calculators with lenders being used currently with some lending providers. Digital verification of identity (VOI), bank statements data scrapped and financially analysed, credit files accessed to all speed up the process of giving you a quick and reasonably accurate financial description of “you”. But what if you went to 4 banks online and did this before choosing which lender to apply with? If they did 4 credit enquiries (1 each) on you this could lower your credit score. As multiple quick enquiries is seen as a negative. What if you have been on holidays in the last 3 months and your expenses are higher than normal for that period? The expenses captured from the bank statements will say you have a higher cost of living than in reality- lowering your borrowing capacity. These are the reasons why we have people to review your information and ask questions
No doubt I will write more on this topic. But at this stage I feel the best approach is to use a balance of tech and human resources. Exciting times really…… more soon
Side note -The irony of this blog is I could use AI to write it in this website platform and probably have it rank better SEO wise etc. Thanks for reading, Luke
