5 mins read
The Financial Services Compensation Scheme (FSCS) provides crucial protection to customers of failed financial services firms. We designed a solution which would allow FSCS to process high volumes of complex claims quickly and efficiently, providing a better service to customers, many of whom were vulnerable.
FSCS exists to protect customers of financial services firms who have gone out of business. If a firm fails, FSCS can step in and pay compensation where customers have lost money, up to a compensation limit subject to certain criteria being met. The claims FSCS sees are becoming increasingly more complex, and each claim has to be assessed under the rules set by the UK regulators, the Financial Conduct Authority (FCA) and Prudential Regulation Authority (PRA).
When a claim is submitted to FSCS, there is, necessarily, an extensive paper trail as customers must provide evidence to support their case this can involve hundreds of documents for a single claim, with some of these each a few hundred pages long. It’s an arduous and time-consuming task for the claims handlers to review all the documentation in order to pull out relevant information.
Our customer management division has been helping FSCS manage surges in claims when a firm fails, so we’ve experienced first-hand how time-consuming and detailed the process can be, as well as witnessing the distress of customers waiting to find out if FSCS can pay them compensation, often life changing sums of money.
With our technological expertise and extensive business processing experience, we were confident we could streamline the process, so we commissioned our strategic consulting arm – Capita Consulting – to find a solution. Here’s what we achieved.
The pilot: using AI to convert 700k voice messages to referenceable evidence
When a large investment fund collapsed, FSCS needed to process huge numbers of claims and get decisions to customers as quickly as possible including having to listen to and analyse more than 700,000 voice messages. We saw an opportunity to build a custom speech-to-text solution using Microsoft’s Azure Speech Services alongside our own tools.
Using a suite of key words and phrases identified by FSCS, we designed the solution to enable the recordings to be transcribed quickly and accurately, and to be saved into an easy-to-use, searchable and referenceable database. After passing rigorous testing and checking by FSCS to ensure a high level of accuracy and compliance with FSCS rules, FSCS were able to use the solution to unlock the insights within the voice messages to support decision-making on individual claims.
By using natural language processing, our pilot solution compressed the human process from 8 months to 3 weeks, providing a significantly more accurate AI model than anything existing in the financial services sector.
Microsoft were so impressed by what we’d achieved with the speech software, just two days before their global Build 2020 event they adapted their agenda to showcase our solution as a key case study.
The next stage: a platform to process high volumes of unstructured data from all sources
As part of our commitment to continuously innovate, we decided to invest our own resources to enhance the customer service we provide to FSCS by building a next-generation platform Accelerate which would process large volumes of unstructured information from any source or format.
As well as leveraging Microsoft’s voice recognition within the platform, we incorporated optical character recognition (OCR), enabling emails, electronic documents, voice messages and even handwritten information to be accurately transcribed and categorised.
Our focus was on ensuring a streamlined, intuitive experience for claims handlers to support them with the initial assessment the triage of claims whilst also offering the ability to search and drill down into documents to further support more efficient claims handling.
The result: transforming the experience of claims handlers, FSCS and their customers
We applied Accelerate to large numbers of claims resulting from a defaulting financial services firm, with impressive results:
- Saving of 1,267 claim handling hours when triaging 700 claims (327 working days reduced to 158 working days)
- Major improvement in triage quality, particularly when many documents are involved in the consideration of the claim
- 80% reduction in costs relating to the triage element of the process
The benefits can be evidenced across all touchpoints of the process, from the customer seeking compensation, to the claims handlers, to FSCS.
Supporting faster, informed claims handling whilst reducing the cost of service
By introducing the latest technologies, Accelerate transforms the process of claims handling when high volumes of documents are involved, significantly speeding up transaction times and therefore customer service response times offering a transformed experience for both customers and claims handlers. This increased efficiency means Accelerate significantly reduces the cost of processing the claims where it’s used.
Minimising a period of great anxiety for customers
People often have their life savings tied up in the investments they’re seeking FSCS compensation for, but it can take some time to find out what’s going to happen to the money they invested. In the meantime they’re vulnerable often experiencing huge stress, worry and significant financial difficulty. By accelerating the process and improving accuracy, our solution shortens that waiting period - decisions can be made more quickly and customers can get back on track as soon as possible.
Sarah Marin summarises the impact of Accelerate: 'We’re committed to making decisions on claims as quickly as possible and, in the case of a recent defaulting firm, we were able to start paying compensation sooner than we would otherwise have been able to without comprising quality. Pioneering the use of artificial intelligence in partnership with Capita is enabling us to improve the experience and outcomes for our customers during a particularly difficult and anxious time in their lives.’