The customer service and experience space is overrun with AI solutions, but how many are designed to tackle real contact center issues?
If you adopt a ‘one-size-fits-all' approach to implementing AI within your contact center, it is likely that you will fail to maximize the potential of the technology.
Generic AI is not suited for areas like contact center audio, due to issues with background noise, accent variation, and overlaps. Standard tools often struggle to adapt to real customer service workflows, leading to too many low-value and time-consuming tasks for agents.
In an exclusive conversation with CX Today, Jonathan Foureur, Head of AI at Diabolocom, provided further insights into some of the shortcomings of generic AI solutions.
AI Developed with Field and Business Expertise
Foureur outlined the fact that many benchmark datasets don’t reflect real-world CCaaS scenarios. He explained that most speech recognition models, for example, are trained on YouTube videos or other high-quality audio sources, which don’t account for noisy call center environments.
To combat these issues, Diabolocom’s AI solutions are purpose-built for the contact center environment, enabling them to understand noisy audio, emotional tone, silence, and escalation signals.
The vendor also collaborates directly with its customers by organizing regular workshops. During the sessions, Diabolocom uncovers the issues that are impacting contact center professionals in their day-to-day roles and co-develops AI solutions that truly meet real-world expectations.
Additionally, to help build on the firm’s track record of producing AI tools that truly address CX-specific challenges, the company recently launched Diabolocom Research. This dedicated research lab is designed to test and create advanced customer service and experience AI solutions.
“The point of Diabolocom Research is to use a scientific point of view to close the gap between the current market conditions and what will benefit the CX space,” Foureur explained.
“We’ve developed open-source datasets tailored to CCaaS use cases and are contributing to the AI research community by addressing industry-specific challenges.”
“This ensures our AI models are trained on data that truly reflects customer interactions, improving their real-world performance.”
The Power of Sovereign AI
Diabolocom Research is led by Kevin El Haddad, Head of AI R&D at Diabolocom and an Applied Machine Learning Researcher at ISIA Lab at the University of Mons, France.
Haddad is supported by a team of PHDs and experienced AI Researchers, who combine their skills to craft cutting-edge technologies across the following focus areas:
- Conversational AI
- Voice AI
- NLP
- Automation
For Diabolocom’s CEO Frédéric Durand, the launch of the research team represents a “significant milestone for us [Diabolocom] as we expand our capabilities to create more intelligent, adaptive, and scalable solutions for our clients.
“This initiative underscores our dedication to staying at the forefront of innovation and AI enhancements to customer experience.”
As well as allowing Diabolocom to stay abreast of the latest trends and developments in the contact center AI space, Foureur also discussed some additional benefits of in-house AI compared to third-party AI solutions.
While third-party providers can often seem effective at demo or prototype level, many organizations make the mistake of not considering scalability, resilience, and actual business impact.
Diabolocom’s AI solutions are designed with the contact center in mind, aiming to support millions of daily interactions efficiently.
Many third-party providers are not robust enough to handle this level of traffic. In addition, they often do not offer service level agreements (SLAs), which means businesses have no guarantees on uptime or performance.
