Mission Lane: Delivering next-level operations through AI

We speak to Mike Lempner, Head of Engineering and Technology at Mission Lane, to delve into ways the company is leveraging AI to streamline its operations, offer greater credit decisioning, underwriting, marketing and fraud solutions, as well as enhance its customer services. 

Can you please provide background on Mission Lane and your role as Head of Engineering and Technology?

Mission Lane is a financial technology company that provides credit products to a diverse range of customers who are typically denied quality financial services due to having poor or limited credit history. 

Using alternative and traditional data, we underwrite individuals based on a variety of factors and develop tailored solutions to help our customers at any stage of their financial journey. 

I lead a team of technologists and software engineers tasked with building the technology solutions we use for credit decisioning, fraud, marketing, and digitally servicing our customers across multiple channels, as well as other functions to support our business. 

How is Mission Lane implementing emerging technologies, such as AI, into its day-to-day business operations?

At Mission Lane, we have used AI, which includes Machine Learning models and algorithms, in several ways since our inception, with the goal of enhancing our credit decisioning and underwriting, marketing, fraud, and operations. 

Depending upon the use case, we may select a particular type of model based on its performance as well as the explainability requirements. 

For example, credit underwriting models need to be explainable to demonstrate that there is no bias or disparate impact from the decisions. 

Other types of operational, fraud, and marketing use cases may not have the same requirements and may be able to benefit from other types of models. 

Generative AI tools like ChatGPT can also be immensely helpful in getting the code-writing process started and generating ideas. 

Recently, for example, we were using a tool that tracks things like software delivery, as well as active tasks and, how long it took to complete them. We wanted to load this data into another system to analyse it, so we leveraged ChatGPT to write the code to do this. 

Of course, we needed to double-check and make some changes to the code, but a process that would normally take hours took minutes and was mostly accurate, providing an upside for productivity. 

We’ve also used Generative AI for visualizing organizational changes. 

One of our developers took a list of everyone in the organization and used Generative AI to generate a website with data broken out by department and manager to model different scenarios, that even incorporated our brand guidelines into the web design. 

This is another task that took minutes to complete with AI vs. manually.

Has the use of AI impacted your customer services?

Our ability to automate tasks that would typically take hours to complete manually has freed up our time to focus more on our customers, delivering quality services, and working with them to improve their financial standing. 

AI can also help us better understand our customer base and analyse trends with their behavior in an efficient manner, allowing us to better serve them. 

For instance, we have used Generative AI to write an SQL (Structured Query Language) that could identify specific types of customers and arrange them in a table. 

This helps us improve overall organisation and better visualise our customer base, allowing us to make informed decisions more efficiently. 

In terms of underwriting, we can’t ask Generative AI solutions to decide if we should approve an individual based on their credit information because we need to be able to justify our decisions and demonstrate how we arrived at a given decision. 

However, these technologies can identify areas for improvement and allow us to gather the information we use more efficiently.  

How do you safeguard against inherent risks of AI, such as fraud and bias?

Ensuring that we are not introducing bias into decision-making requires diligence on our end, and not only closely analysing the outputs but also double-checking the inputs in our models. 

AI/ML may process the data and predict outcomes, but data scientists still need to justify and explain underwriting decisions to actual humans.

While AI can be beneficial within organisations, it can also be leveraged by bad actors outside the organisation. 

Identity fraud is a leading concern given the potential of AI to mimic an individual’s voice or generate doctored images. Know Your Customer (KYC) checks are essential to ensuring data and finances are protected.  

That said, AI can also be used to help organizations identify fraud and flag suspicious activity that could be missed. 

We leverage these capabilities using internal data, but also rely on fraud vendors who may be leveraging their own AI/ML capabilities and data sources.

What are some best practices for educating employees on the implementation of AI? Is there any advice you could provide to technology leaders about onboarding teams to the technology? 

My advice for IT leaders at fintechs is to provide comprehensive education and have regular touchpoints with colleagues to explain the technology in layman’s terms. 

Clear and consistent communication on how AI is being used and its benefits for business is essential for acceptance and adoption. At the same time, employees need to be educated about how and when it’s appropriate to use Generative AI tools like ChatGPT (and when it’s not). 

For example, companies may want to avoid their employees inputting sensitive or proprietary data. However, for certain use cases, it may be necessary to leverage Generative AI solutions running internally that are trained on and leverage internal data. 

What regulations can we expect around AI for the fintech industry? 

While it’s still early in the AI revolution, we expect most regulations around AI to surround data transparency and explainability and aim to eliminate bias and justify how decisions are made and what data is being used. 

We also expect regulators to increasingly invest in technology that can determine whether content or a decision was made by AI and set disclosure requirements and limitations on what AI can be used for. 

This is an evolving landscape and organisations should be nimble in ensuring compliance. 

What impact do you foresee AI having on financial services for the underbanked?

At least at Mission Lane, the biggest impact we see AI having on our customer base is to enable us to continue to enhance our models to be able to offer credit to more underbanked customers. 

With Generative AI, there are also opportunities to automate many more manual tasks to free up our employees’ time for more critical activities. 

Instead of spending hours writing code or sifting through data, we can spend the time providing better, tailored services with personalised touchpoints. 

We can devote more time to brainstorming creative solutions, identifying unmet needs, and improving the user experience.

******

For more insights from FinTech Magazine, you can see our latest edition of FinTech Magazine here, or you can follow us on LinkedIn and Twitter.

You may also be interested in our sister site, InsurTech Digital, which you can also follow on LinkedIn and Twitter.

Please also take a look at our upcoming virtual event, FinTech LIVE London, coming on 8-9 November 2023.

******

BizClik is a global provider of B2B digital media platforms that provides executive communities for CEOs, CFOs, CMOs, Sustainability Leaders, Procurement & Supply Chain Leaders, Technology & AI Leaders, Cyber Leaders, FinTech & InsurTech Leaders as well as covering industries such as Manufacturing, Mining, Energy, EV, Construction, Healthcare, and Food & Drink.

BizClik – based in London, Dubai, and New York – offers services such as Content Creation, Advertising & Sponsorship Solutions, Webinars & Events.