ARM
Nominated Award: Best Application of AI to achieve Social Good
Website of Company: https://www.armonitoring.co.uk/
We have a proprietary AI model that is used by financial firms and utility companies to identify vulnerable customers. The model uses financial and banking transactional data (open banking) to identify vulnerable customers and the characteristics of vulnerability they display. Our AI model has evaluated 2,5 million individuals and was selected by the FCA and the City of London Corporation for their Digital Sandbox Pilot. We have a partnership with an online banking platform to provide our reports to their corporate customers (lenders and utility companies).
It is estimated that almost 50% of the adult population displays at least one characteristic of vulnerability.
Every day vulnerable customers are impacted in their daily lives by customer service failures. More than 30% of the adult population struggle to understand their monthly mortgage statement information, 50% cannot afford an unexpected expense of 300 euros and more than 15% of the population has no internet access to deal with banking digitalization.
Vulnerable customers have different needs depending on their characteristics and primary services firms like utilities or financial firms are slow to manually identify every vulnerable customer in their portfolio, failing to stay ahead of issues. In 2020, just Barclays treated without the appropriate levels of care 1.5 million customers in financial difficulties.
ARM helps financial firms prevent and remediate any issue before the customer suffers any inconvenience.
ARM has designed a solution that provides the first vulnerability passport to customers in difficulties so they can share that information with their electricity, gas, water or financial provider to get the appropriate service and products. ARM solution identifies early signs of vulnerabilities using Data Analytics and Machine Learning. ARM then classifies vulnerable customers according to their characteristics and provides that information to utilities and lenders to prevent any harm to these customers.
To identify early signs of vulnerability, we use data analytics and a proprietary machine learning model to review transactional data with high accuracy (>90%) and precision, which improves with every new customer it evaluates. We’ve added a layer of explainability to provide a scoring system indicates if a customer is likely to be in challenging circumstances. Our vulnerability passport classifies factors of hardship into: •Financial Difficulties •Health • Accessibility • Life events.
With ARM vulnerable passport, customers can be sure that their utilities and financial providers will take every necessary step to treat them according to their needs in the most important aspects of life.
Reason for Nomination
ARM founders have been working for decades in the financial industry and were tired of the slow and manual quality processes that were slow and prone to error. These methods are reactive and most of the time starts once financial firms receive many complaints and a systemic issue has caused severe harm to a large customer base. Vulnerable customers are the first to notice this harm and suffer the worst consequences.
We are currently working on two use cases, for utility companies and lenders. Still, we aim to expand the use of the vulnerability passport to other industries to protect vulnerable customers (e.g. insurance or banking). Utility companies use our model to evaluate customers in arrears that are about to enter a litigation process to disconnect them from the grid. Several laws and regulations prevent companies to litigate against clients in severe vulnerable circumstances (e.g. a customer displaying health or taking care of elders or youngers). With our explainable AI model, utility companies can identify the customers they shouldn’t litigate against early on, preventing unnecessary stress on vulnerable customers.
Lenders use our AI model to evaluate customers in arrears before an agent contacts them to propose a repayment plan. These agents are young, lack working experience and have poor training. On average, they need an hour to agree on a repayment plan with the borrower in arrears, and very often, those plans fail a few weeks later because the repayment plan was not adequate to the borrower’s circumstances. Lenders use our vulnerability passport, which provides all the information they need before they call the customer, to set up repayment plans that suit the borrowers’ financial and personal circumstances. ARM uses big data and AI/machine learning technology. The model uses cluster analysis to identify patterns shared by potentially vulnerable customers, extracting insights like poor resilience or transient vulnerabilities that otherwise would not get noticed. In addition, ARM solution includes algorithms that identify predefined vulnerability triggers (e.g. consecutive missed payments).
Recently, ARM has been adding additional databases to provide the most complete set of information and will be able to provide more detailed information on Health difficulties, which is one of the most burdensome and difficult vulnerabilities to identify.
Additional Information:
ARM is an innovative Irish Company headquartered in Cork that is currently focused on the UK market but aims to expand to the entire European Union once we prove our benefits to UK financial firms.
ARM has been one of the few European startups selected for AWS FinTech Accelerator 2022.
Our founders are a reference in compliance and AI, acting as panelists in top tech events like Boston’s Fintech Week talking about ‘Bias in AI’.
ARM project 2021
Name of Company: Automated Regulatory Monitoring
Websites of Company: www.armonitoring.co.uk
Nominated Award: Best Application of AI to achieve Social Good
It is estimated that almost 50% of the adult population displays at least one characteristic of vulnerability.
Every day vulnerable customers are impacted in their daily lives by customer service failures. More than 30% of the adult population struggle to understand their monthly mortgage statement information, 50% cannot afford an unexpected expense of 300 euros and more than 15% of the population has no internet access to deal with banking digitalization.
Vulnerable customers have different needs depending on their characteristics and financial firms are slow to manually review every vulnerable customer journey, failing to stay ahead of issues. In 2020, just Barclays treated without the appropriate levels of care 1.5 million customers in financial difficulties.
ARM helps financial firms prevent and remediate any issue before the customer suffers any inconvenience.
ARM has designed a solution that monitors a tailored customer journey for each customer based on her vulnerability characteristics. Financial firms can identify every day if a customer has not been treated with appropriate levels of care at any step of the journey.
Our solution identifies early signs of vulnerabilities using Data Analytics and Machine Learning. Then classifies vulnerable customers according to their characteristics. Finally, we monitor the customer journey based on those characteristics.
To identify early signs of vulnerability, we use data analytics and a proprietary machine learning model to review transactional data with high accuracy (>90%) and precision, which improves with every new customer it evaluates.
We have developed +70 algorithms to monitor the customer journey following more than 12 regulations and guidelines from the FCA, the Money and Mental Health Policy Institute and the Money Advice Liaison Group. Each of these algorithms monitors one key customer service step.
With ARM vulnerable customers can be sure that their financial firm will take every necessary step to treat them according to their needs, preventing any harm from their lives most important financial decisions.
Reason for Nomination:
ARM founders have been working for decades in the financial industry and were tired of the slow and manual quality assurance processes that relied on random samples of a few documents. These reviews are reactive and most of the time start once financial firms receive many complaints and a systemic issue has caused severe harm to a large customer base. Vulnerable customers were the last to notice this harm and suffer the largest monetary loss.
ARM founders wanted to change the paradigm from a reactive to a proactive approach. Using AI and data analytics to oversee a tailored customer journey for 100% of the customer portfolio every day, financial firms can prevent any harm to vulnerable consumers before they suffer any inconvenience.
ARM uses big data and AI/machine learning technology. The model uses cluster analysis to identify patterns shared by potentially vulnerable customers, extracting insights like poor resilience or transient vulnerabilities that otherwise would not get noticed. In addition, ARM solution includes algorithms that identify predefined vulnerability triggers (e.g. consecutive missed payments).
ARM solution is the only flag system tailored to monitor vulnerable customers. It does not require any customization because it uses standardized predefined rules that can survey
any database to identify all types of vulnerabilities and can be easily deployed in 4 weeks, making our model highly scalable and a low-cost solution for firms of any size.
ARM solution was able to evaluate 2.5 million individuals in 24 hours during the FCA digital sandbox pilot, saving 1,25 million hours of compliance specialists to review the entire portfolio, as it would take at least 30 minutes to evaluate each customer and the entire customer journey.
ARM has mapped more than 12 regulations and guidelines that account for more than 1,000 pages to create the features of the ML model and designed hundreds of tailored customer journeys. On top of that, they had to go over thousands of data fields to identify less than 100 commonly used data fields in financial firms’ databases to make the solution scalable. Finally, they had to go through a thorough analysis of the training data sets and results to ensure the model is accurate, precise, and unbiased.
Additional Information:
ARM is an innovative Irish Company headquartered in Cork that is currently focused on the UK market but aims to expand to the entire European Union once we prove our benefits to UK financial firms.
Our founders are a reference in compliance and AI, acting as panelists in top tech events like Boston’s Fintech Week talking about ‘Bias in AI’.
ARM has been recently shortlisted at the Credit & Collections technology awards for the best Vulnerable Customer Identification & Screening solution.