Resolving 13,000 Customer Issues a Month with Conversational AI

• 45,000 interactions per month through the AI front door

• 20-30% containment (≈13,000 resolved without an agent)

• 90% of inbound contact routed through the AI layer

• Reduced AHT through context-rich handovers

• Agents freed from repetitive admin

Client Goal 

Create a modern aftersales support experience that makes it easier for customers to get help while enabling the manufacturer's service teams to manage growing demand without increasing operational cost. 

Challenge 

Rising contact volumes and fragmented service journeys meant customers often had to navigate multiple steps to resolve routine issues such as faults, warranty queries or appointment changes. This increased customer effort and placed growing pressure on frontline teams. 

Solution 

FourNet and the manufacturer designed and deployed a conversational AI "digital front door" across voice and digital channels. Customers can explain issues in natural language, complete common service journeys end-to-end, and escalate seamlessly to an agent when needed. 

Outcomes 

The platform now manages around 45,000 interactions every month, achieving 20-30% containment across channels and resolving more than 13,000 customer issues without agent involvement. 

For appliance manufacturers, the product experience does not end at the point of sale. When something breaks, when an engineer visit needs to change, or when a warranty question arises, the quality of aftersales support becomes the defining moment of the customer relationship. 

For the manufacturer, aftersales support is therefore a critical part of brand experience. The service operation must manage high contact volumes reliably while keeping the journey simple for customers and workable for frontline teams. With the introduction of the EU's Right to Repair Directive in July 2026, there will be more pressure on white goods providers that also sell in the EU. 

That challenge has been intensifying. The manufacturer handles around 45,000 customer interactions every month, many relating to routine but time-sensitive requests such as reporting faults, checking warranty cover, or rearranging engineer appointments. 

The opportunity was clear. If these journeys could be simplified and automated intelligently, the manufacturer could improve the customer experience while also reducing operational pressure. 

When Routine Service Journeys Become Complex 

Before the transformation programme began, many of the most common aftersales interactions followed processes designed around internal systems rather than customer outcomes. 

Customers often needed to move through multiple steps to complete relatively straightforward tasks. Fault reporting might require several stages of questioning. Warranty validation could involve separate checks across different systems. Appointment changes frequently required manual intervention from agents. 

The result was friction on both sides of the interaction. 

Customers sometimes had to repeat information as conversations moved between channels or agents. Meanwhile, frontline teams were spending significant amounts of time handling administrative steps rather than focusing on the complex or emotionally sensitive cases where human support makes the biggest difference. 

Over time, this created three operational pressures. 

Avoidable demand as customers contacted the manufacturer more than once to resolve the same issue.

Repeated data capture where agents needed to re-collect information already provided earlier in the journey. 

Reduced capacity for complex support because large volumes of routine administration consumed valuable agent time. 

Create a Single "Digital Front Door" for Customer Support 

The manufacturer's objective was not simply to introduce automation. The ambition was to redesign the way customers access support. 

The goal was a scalable digital front door that could guide customers to the right outcome quickly, automate routine service journeys end-to-end, and involve human agents only when their expertise was genuinely needed. 

This meant designing an experience that worked naturally for customers while integrating deeply with the operational systems that power aftersales services. 

Conversational AI That Understands Customers in Their Own Words 

FourNet worked with the manufacturer to design a conversational AI experience that allows customers to describe their problem naturally rather than navigating rigid menus. 

The platform supports voice and digital channels, identifying customer intent through natural language understanding and guiding users through the steps required to resolve their issue. 

Common journeys now handled by the platform include: 

• Fault diagnostics 
• Warranty validation  
• Booking engineer visits 
• Amending existing appointments 

Where a situation requires human support, the AI performs a warm transfer. The conversation transcript and a structured summary are passed directly to the agent so the customer does not need to repeat information. 

This approach ensures automation removes friction rather than creating it. 

Where a situation requires human support, the AI performs a warm transfer. The conversation transcript and a structured summary are passed directly to the agent so the customer does not need to repeat information. 

This approach ensures automation removes friction rather than creating it. 

From the beginning, the conversational experience was designed using plain English and clear explanations rather than technical language. 

The assistant explains what information it needs and why, guiding customers through each step to reduce unnecessary back-and-forth. 

Particular attention was also paid to inclusivity. UK customers describe appliance problems in many different ways, and regional accents add further variation. 

To ensure accessibility, the speech recognition and natural language models were trained to recognise different accents, speech patterns and ways of describing appliance faults. 

This allows customers to explain problems in their own words rather than adapting their language to the system. 

The programme followed a phased delivery approach. 

A live proof of concept launched in July 2025, validating the AI's ability to handle complex aftersales journeys such as diagnostics and engineer scheduling. 

Following the success of this phase, the full production rollout was completed by November 2025, supported by two-week hypercare periods at each deployment stage to stabilise performance as interaction volumes increased. 

Today, the conversational AI operates across voice, webchat and WhatsApp channels as the single entry point for aftersales support. 

Behind the conversational experience sits a tightly governed technical architecture. 

The solution is orchestrated using Druid AI and integrates Azure OpenAI GPT-4o mini, hosted within the Azure EU environment to support data residency and GDPR compliance. 

The AI connects in real time with multiple operational systems including: 

• CRM platforms 
• Booking and scheduling systems 
• Warranty validation services 
• Address verification tools 
• Telephony infrastructure 

All integrations use secure REST and event-based patterns, ensuring the assistant can complete tasks rather than simply providing information. 

Governance was embedded throughout delivery, with oversight from the manufacturer's Data & Information Security Officer and strict prompt guardrails to ensure the assistant operates within defined service boundaries.

Faster, Simpler Support for Customers 

Customers can now describe their problem in natural language and reach the right service path more quickly. 

When escalation is required, agents receive the full context of the conversation, reducing repetition and shortening the time to resolution. 

Operational Efficiency at Scale 

The digital front door now manages approximately 45,000 customer interactions every month. 

Across channels, the platform achieves 20-30% containment, meaning more than 13,000 interactions are resolved entirely by the AI each month. 

This reduces pressure on contact centre teams while ensuring customers can still access human support when needed. 

Better Use of Agent Expertise 

A core design principle was to remove repetitive administrative work from the agent role. 

By automating high-volume routine journeys and passing structured context when escalation occurs, agents now spend more time handling complex or emotionally sensitive situations. 

This improves both the quality of customer conversations and the sustainability of the workload for service teams. 

FourNet's role in the programme extended beyond implementing technology. 

The team worked with the manufacturer to redesign the aftersales journey itself, identifying where friction existed and where automation could genuinely resolve customer needs end-to-end. 

This approach combined: 

Customer experience design to simplify journeys and reduce avoidable demand.
Enterprise-grade AI integration to connect conversational interfaces directly to operational systems.
Operational delivery discipline to ensure the platform performs reliably under real-world service conditions. 

For the manufacturer, the digital front door represents more than a new contact channel. It establishes a scalable service model that can evolve alongside customer expectations. 

By combining conversational AI with strong operational design and governance, the manufacturer has created an aftersales platform capable of handling high volumes efficiently while maintaining a human, supportive experience for customers.