Using AI to help make meaningful decisions
Designed to explore how AI could enhance large data set interpretation by replacing manually defined business rules with data-driven decisioning.
Working closely with our financial services clients we explored how we can use AI to demonstrate the value of our patent-pending Fraud Optimised Decision Engine (ODE), designed to transform fraud detection in financial services. The goal was to show how ODE could enhance manually defined rulesets with auto-generated and optimised rules tailored to real-world scenarios.
We developed a proof of value to demonstrate how ODE could significantly enhance fraud detection, with the potential to reduce financial risk and strengthen regulatory compliance. ODE successfully demonstrated improved fraud detection performance and delivered business-aligned results by maximising true positive detection whilst minimising unnecessary false positive interventions.
Using AI to make our daily jobs more efficient
Generative AI (GenAI) is being used within one of our key accounts to accelerate project delivery. The client has an approved GenAI solution that allows the Sopra Steria team to more quickly write and check the quality of code, which expedites the delivery of projects in support of the client's core mission.
We are also exploring using GenAI to write test cases to increase test coverage. Importantly, all AI‑generated code will be strictly controlled and carefully reviewed by both our specialist staff and scanning tools to ensure it doesn’t introduce errors or vulnerabilities. By combining the efficiency of GenAI with the expertise of our people, this will help to increase our code coverage, improve code quality and reduce the time to market.
Using AI to help us work smarter not harder
Clever use of AI within our business means we can help support our people to work smarter and more efficiently.
To further our use of AI within our contact centres, we will be introducing AI-enabled call handling for our Amazon Connect and Salesforce platforms, making life even easier for our contact centre agents and the customer experience even better by helping to resolve queries more quickly.
Using AI, the platform will suggest the next best action to the call handling agent based on similar case resolution history. During the call, the voice interaction will automatically be converted to text avoiding any typing by the agent for super efficiency. At the end of the call the AI will suggest a summary note for the case which the agent can choose to accept removing the need for the agent to type this in manually. The customer will see this on the portal too. This saves time and effort for everyone involved in the call – benefiting everybody.
We are continuously seeking new ways of smarter working with AI to use within our business using the technologies we have. This includes a new self-serve portal powered by AI available through our Salesforce platform for our NHS Shared Business Services customers to help answer their queries anytime – day or night. The portal aims to provide answers to questions raised there and then, without directing the customer off to seek information elsewhere. This is about enhancing the customer interaction by making the engagement with the agent quicker and easier and much more convenient too.
Our focus for AI is not to replace but support our people and enhance their work using these AI based technologies. AI is being used to enhance our information retrieval process to quickly and accurately find time-sensitive data.
AI-enabled data retrieval and analysis
Our project management and service management professionals now have their own dedicated AI tool. This tool has been developed by Sopra Steria’s in-house experts and takes advantage of Open AI’s GPT4’s Large Language Model. The tool runs on our own infrastructure and is insulated from the internet with in-built security and privacy mechanisms. Using the tool enables our project and service managers to cross-examine project specific documentation to find and analyse information more quickly and easily.
The tool can:
- Review key documents and highlight risks to delivery.
- Summarise key obligations and dependencies.
- Interrogate key company-wide policies, guidelines, and best practices faster than traditional mechanisms.
We are dedicated to the careful and responsible use of AI. To ensure robust oversight, this tool has undergone review by our AI Governance Board, with the appropriate controls and standards in place. Additionally, we conduct account-level assessments to determine whether AI can be appropriately deployed within each context.
We’re already seeing benefits around new insights and our staff are finding it exciting to have a practical application of AI in the project and service management space.
Enhancing our contact centre interactions by using AI
We are using AI agents to help us process high volume frequently asked questions (FAQs) within our contact centres, freeing up more time for our colleagues to have valuable in-depth human-to-human interactions and conversations with our clients.
We have implemented noise reduction technology in our contact centres enabling us to have clearer conversations with our clients by significantly reducing background noise. This technology is invisible, working behind the scenes to suppress noise on both ends of the call, especially when working environments are often busy and noisy. Both contact centre staff and clients can focus on the content of the conversation without having to worry about background noise levels and interruptions.
The graphic below shows the advisor noise impact with Iris Clarity represented as the green line and the associated headline improvement statistics:
AI Agents in Action
We are using AI in tandem with our people to provide higher quality services to our clients in our contact centres by using AI powered agents across both our voice (internally known as Luna – conversational AI) and chat (text based) channels (internally known as ORA – On-demand Response Assistant) as the first line of communication for FAQs.
In practical terms this means those typical questions such as ‘When is payroll cut-off?’ will be answered more efficiently for our clients, while reserving our highly trained colleagues for those more complex queries such as ‘I've not been paid’, to be available and the call to be passed quickly through to the next available agent.
For Luna, where we have been able to answer a topical question and the customer still would like to speak to an agent, we can route them straight through – rather than typically presented Interactive Voice Recognition (IVR) options enhancing the customer experience.
For ORA and our text-based chat services, ID and verification checks are performed at the initial engagement with any customer at the start of any chat. This means that at any point a customer is transferred to an agent, all the ID and verification data is passed through - including relevant details and the original question asked avoiding further verification checks and improving the overall customer experience. This benefits the customer by creating a better experience, but also the contact centre advisor because the checks have already been completed and passed through, saving valuable time.
The human touch will always be required for those calls that require compassion or confidential human resources matters – so deploying AI agents to manage those straightforward requests makes sense and frees up resources to handle those more sensitive matters.
Noise Reduction Technology
We are using AI agents to help us process high volume frequently asked questions (FAQs) within our contact centres, freeing up more time for our colleagues to have valuable in-depth human-to-human interactions and conversations with our clients.
We have implemented noise reduction technology in our contact centres enabling us to have clearer conversations with our clients by significantly reducing background noise. This technology is invisible, working behind the scenes to suppress noise on both ends of the call, especially when working environments are often busy and noisy. Both contact centre staff and clients can focus on the content of the conversation without having to worry about background noise levels and interruptions.
The inverse that a noisy environment is noticeable and distracting. Before implementing this technology, our customer surveys often had comments about the high levels of background noise – which have now faded away since introducing this AI-powered technology.
Benefits include improved employee wellbeing from a reduction in stress due to a quieter, more focused environment where employees are likely to feel more satisfied due to effective communication with less distractions. We have found:
- Improved customer satisfaction survey results were 89% before Iris Clarity was introduced, rising to 93% post-implementation.
- Internal advisor satisfaction increases through clearer communication and less distraction.
- High quality audio is often associated with professionalism improving our overall brand image.
"The deployment of IRIS Clarity marks a significant step forward in enhancing our Customer Experience Management Hub. By eliminating background noise, we’re enabling clearer conversations, reducing agent stress, and improving customer satisfaction—making every interaction more efficient and professional." Matt Speight – Product Owner, Sopra Steria
"With IRIS Clarity, I no longer need to repeat myself or strain to hear what customers are saying through background noise. My calls now flow more easily and allow me to concentrate fully on providing excellent service." Ricky – CXM Hubs, Sopra Steria
Governing AI activities for our business
At Sopra Steria we like to do things both ethically and responsibly, which is why we set up our own internal AI Governance Board (AIGB), providing a set of standards and controls across all our AI initiatives.
The AIGB crucially reviews and approves all AI activities in a streamlined way, ensuring that we are doing the right things, at the right time, in the right way. Our work is grounded in Sopra Steria Group’s AI principles that are closely aligned with the UK Government’s ten principles for the use of AI.
All AI activities are coordinated through the board providing our senior leadership with a centralised view of all AI activities happening within our organisation. Detailed AI activity reporting helps our leadership team navigate ethical, legal and commercial considerations.
Through the AIGB, training has been developed to provide all our Sopra Steria employees the opportunity to engage and learn about the new AI tools we have implemented, how these impact our environment, and how we can best use them to make our jobs easier.