Avoiding an ‘us too’ approach to AI

Solving genuine problems using Artificial Intelligence

Does a fridge really need AI?

There’s a product launch almost every day boasting a new spin on Artificial Intelligence, and when even white goods manufacturers are getting involved it starts to feel like we’ve hit ‘peak AI’.

AI shouldn’t be a marketing gimmick used to pique people’s interest. It should be a tool to solve genuine problems — problems that it would be impractical or impossible for a human to tackle.

AI is at the core of what we do at Mojo.

Out with the old

We’re doing away with the bottlenecks that make traditional mortgage applications such a chore.

From the ping-pong game of returning missed phone calls to the chasing for updates and attending appointments — it’s hugely inefficient.

Digitising the process and retaining these bottlenecks would be self-defeating. Instead, we use AI to make the mortgage application efficient in a way that humans alone simply can’t.

In with the Nu’

For example, a mortgage broker staffed by humans alone can’t be ‘always-on’. In a world where people expect to transact on their mobiles at any time, this is a genuine problem.

Our solution is an AI chatbot that uses Natural Language Understanding (NLU) and Natural Language Processing (NLP).

Through NLU and NLP, the bot will understand, process and answer customer questions at any time. Where it can’t understand a question, we can have an expert human adviser seamlessly interject to help.

Ringing the changes

We have a duty of care to our customers. Part of this involves calling them to confirm some of their application details and make sure they have all the information they need.

AI helps us make these conversations more efficient for everyone involved.

Our dialler platform will analyse the information we have about our customers to recommend times when they are most likely to answer the phone.

This means we won’t hassle customers at inconvenient times and negatively affect the way they think about us. It will also make our communication strategy more efficient by reducing the man-hours needed to help customers through our journey.

A human-bot team

Our algorithm determines the most appropriate products for customers. As it continues to process data going forward, it’ll become a prime candidate for machine learning.

Blending our customers’ responses with third-party data will create a self-teaching algorithm that will continue to evolve, returning ever-more accurate recommendations.

Crucially, the algorithm has an interdependent relationship with our mortgage advisers.

Their expertise helps train the algorithm and, with the advent of Open Banking, it will help them to provide the best advice.

For example, a customer can tell us about themselves to an extent, but they won’t know how their spending patterns affect their chances of getting a mortgage.

As we integrate with Open Banking, the algorithm will support our advisers to offer more tailored advice than ever before by processing customers’ income/outgoings data in a way humans can’t.

Agnostic AI

If we wanted to, we could tweak our algorithm to favour lenders who pay us more commission. We don’t, and we never will.

In theory, a broker’s advice could be also swayed (perhaps even unconsciously) by not wanting to deal with a lender they find difficult, or by having a personal relationship with another. The beauty of using AI is that it can be agnostic, and only return the best result for the customer.

The future

We’ll use AI to help people understand whether the asking price for a property is fair. By analysing historical prices of lookalike properties we can provide insights that really help people — especially if they’re new to the market.

In marketing, AI will help us communicate in a more sophisticated way and retain customers — only using the data they give us to ensure we never send them something irrelevant.

Better outcomes

For Mojo Co-founder and CEO Richard Hayes, AI is key to delivering a service for customers that hasn’t been possible until now.

Our approach to AI solves genuine customer problems that make the traditional brokerage model so flawed.

Richard said: “AI is creating truly better outcomes for customers — saving them money and getting them the best possible advice.”

This story is published in The Startup, Medium’s largest entrepreneurship publication followed by 329,974+ people.

Subscribe to receive our top stories here.