Future Labs

002

In research

Aura

Hyper-personalised customer journeys.

An AI customer companion for organisations with long-running customer relationships. It answers routine requests instantly, spots early signs that someone may be struggling, and hands complex or sensitive cases to the right person.

The question

Can an AI companion handle routine customer requests and spot difficulty early, so an organisation can grow its customer base without growing its service team at the same rate?

Built with

Planned: Flutter app for mobile and web, ASP.NET Core, Azure OpenAI, Azure Machine Learning, Azure SQL. Hosted in Azure.

Status

In research

The problem

In lending, energy, insurance and other services where customers pay over months or years, most contact is routine: what do I owe, can I move my payment date, can I have a statement. Each request still joins a queue, so service teams have to grow in step with the customer base.

The signs that a customer is struggling usually surface late. An organisation often first learns of difficulty when a payment is missed, by which point the options are narrower and the conversation is harder for everyone.

In regulated sectors the bar keeps rising. Under rules such as the FCA’s Consumer Duty, firms are expected to support customers in vulnerable circumstances and to evidence good outcomes. Reviewing calls and emails by hand to prove it is slow and patchy.

What we built

Aura is at the research and design stage. What exists today is a clickable concept that shows how it would look and feel; development hasn’t started yet.

The idea is a personalised app and web experience between an organisation and its customers, bringing together four things: a clear view of where the customer stands, such as balance, payments and what’s next; instant self-service for routine requests like settlement figures, payment date changes and statements, explained in plain English; a safety net that notices early signs of difficulty and offers help before a payment is missed; and timely personal insights, such as when paying a little early would save money.

An AI assistant answers questions in conversation and hands over to a person, with the full context, whenever a case is complex or sensitive.

How we built it

Aura is designed to sit on top of an organisation’s existing systems, not replace them. The core platform stays the source of truth: Aura reads customer and account data through an integration layer, adds AI-generated support and insight, and sends actions such as payments or date changes back for the core system to process.

The planned integration layer connects through modern APIs, older web services, direct database access or scheduled file transfers, so it can work with legacy platforms as well as new ones. The planned stack is a Flutter app for mobile and web, an ASP.NET Core back end, Azure OpenAI for the conversational assistant and written insights, and Azure Machine Learning for predictive models, all hosted in Azure.

Why we built it this way

Integration first. Most organisations we work with can’t replace a core platform to improve customer experience, so Aura has to add value without touching it.

Self-service first, people for what matters. The aim isn’t to remove people from customer service. It’s to take the routine volume off their desks so specialists spend their time on complex and sensitive cases.

Help before arrears, not after. We want to test whether signals from how customers use the app, and what they tell the assistant, can flag difficulty earlier than a missed payment does, and route those customers to support rather than collections.

What we found

Nothing to report yet: Aura is in research. Our early work has mapped where service teams spend their time and which requests come up again and again, such as settlement figures, payment changes, documents and hardship support.

That has shaped the order we plan to build in: instant settlement figures first, then a hardship and affordability assistant, then documents generated on demand.

As with every Future Labs project, we’ll publish what we learn, including what doesn’t work.

Safety and governance

Aura is designed so that people stay in control of anything that matters. Decisions with real consequences for a customer, such as a hardship arrangement, would follow the organisation’s own rules, with specialists able to review and refine the AI’s suggestions.

Every interaction would be logged to give a complete audit trail. Customer data would stay in the organisation’s own secure cloud environment and would not be used to train public AI models.

Spotting financial difficulty is sensitive. How and when Aura uses those signals, and how customers are told, is a core part of the research.

From lab to live

Aura is a concept for any organisation with an ongoing customer relationship and regular payments. These are some of the situations we’re exploring.

A lender. A borrower wants to know what it would cost to settle early. Instead of calling and waiting for a letter, they get an exact figure in the app in seconds, explained in plain English, and can act on it there and then.

An energy supplier. A customer’s payments start arriving later each month, and they ask the assistant about lowering their direct debit. Aura recognises the pattern, offers to move the payment date to payday and points them to support before they fall behind.

An insurer or membership business. As a renewal approaches, Aura shows the customer what they pay, what has changed and the options open to them, and passes anyone who seems unsure or distressed to a person.

We expect to start with a focused pilot on one customer journey, connected to the client’s existing systems.

Future Labs

Common questions

Is Aura a replacement for a customer service team?

No. Aura is designed to take routine requests off the queue, such as balances, settlement figures and payment changes, so your team can focus on complex and sensitive cases. Customers can reach a person whenever they need one, and the assistant hands over with the full context.

Would we need to replace our core systems to use Aura?

No. Aura is designed to sit on top of your existing platform, which stays the source of truth. It reads data through an integration layer and sends actions back for your system to process.

Is Aura available now?

Not yet. Aura is at the research and design stage and development hasn’t started. If the problems it tackles sound familiar, we’d like to hear from you: early conversations shape what we build.

Ready to turn frontier tech into reality?

Speak directly with our team in Edinburgh or Glasgow about your platform, data or AI initiative.

Georgina Hart, Head of Brand & Strategic Growth at ClearSky

Georgina Hart

Head of Brand & Strategic Growth

georgina.hart@weareclearsky.comEmail Georgina

Explore other Future Labs products

See what else we’re building and testing in Future Labs, our self-funded R&D.

Explore Future Labs
Sentinel, a ClearSky Future Labs prototype, shown as a floating 3D object

#003

In development

Sentinel

Human-in-the-loop telemetry.

Explore Sentinel

Aura, ClearSky's personalisation prototype, shown as a floating 3D object

#002

In research

Aura

Hyper-personalised customer journeys.

Explore Aura

Polaris View, ClearSky's AI decision prototype, shown as a floating 3D object

#001

Live

Polaris View

Better, faster, human-led decisions.

Explore Polaris