Artificial intelligence is dominating headlines and public chatter. The explosion in popularity of large language models like ChatGPT has catalysed excitement about AI’s potential to transform the way we work.

One type of AI that will radically transform the way businesses work is Conversation AI, which uses natural language processing to understand and interpret conversations from anywhere in a business, from meetings to phone calls to emails and SMS.

With the ability to analyse these conversations and provide valuable insights for businesses, Conversation AI has many incredible uses for the workplace. Gartner anticipates that by 2025, 75% of all business calls will be captured for data analytics.

Embracing Conversation Intelligence will help organisations manage large amounts of conversations and their data in much more efficient ways. AI can write meeting notes, detect sentiment, flag complaints and provide analysis on conversational data to support customer experience, employee wellbeing, sales performance and more. For individuals, it can eliminate to-do lists. For corporations, it can reveal trends and deliver real business benefits.

The Evolution of Conversation AI

While many of the latest AI developments seem as if they’ve just arrived on the scene, the evolution of conversation AI has been decades in the making. As early as the 1960s, the first chatbot ELIZA was created, though capabilities were limited. Since then, major milestones have progressed Conversation Intelligence, such as breakthroughs in the 1990s in automatic speech transcription, laying the foundation for today’s advanced contextual comprehension. In 2012, deep learning allowed AI models to analyse the nuances of meaning within conversations, rather than just recognising keywords. Today’s best Conversation Intelligence products now intelligently listen and understand the context of a conversation in real-time, and the importance of certain topics and discussion, making them highly effective for unlocking insights from conversation data.

Conversation AI writes your meeting actions for you

Now, Conversation AI is set to revolutionise the ways we work and the ways businesses operate. Conversation AI solutions like Dubber’s ‘Moments’ leverage cutting-edge AI and Natural Language Processing (NLP) techniques to automatically detect pivotal moments in voice, video, chat and SMS data across networks. This type of technology transcends basic keyword searching AI capabilities to deliver nuanced Conversation Intelligence by pinpointing critical snippets, context, sentiment and interactions within conversations. This helps businesses overcome the challenges of navigating vast amounts of conversation data by transforming it into actionable intelligence.

This Conversation AI can detect commitments made during a conversation, such as over the phone, and automatically convert these into tasks and reminders for an employee. As such, this AI facilitates an overhaul of productivity, helping you stay on top of your work, reminding you of those little, easily-forgotten commitments and practically eliminating the need for a to-do list by writing your meeting actions for you, whether that meeting is over a UC Platform such as MS Teams, Webex, or an on-the-go phone call.

Understanding sentiment: Wellness and Customer Service

The best Conversation Intelligence technology doesn’t only detect commitments and actions made in conversations, but can also analyse the context and sentiment of the language used. For example, if a customer is issuing a complaint over the phone, or an act of verbal abuse occurs, these moments of negative sentiment and complaint are detected by the AI and can be proactively flagged. For Customer Service, this means that the most pressing grievances can be addressed as a priority, or resources can be reshuffled to resolve a bottleneck, improve practice and future training and ultimately resolve to boost employee morale.

One industry where this can be highly effective is local government. For example, Dubber’s ‘Moments’ was deployed by the Devonport City Council (DCC) call centre to help facilitate improved operational efficiency for the support services offered by the council to its community of 26,000. In consultation with Telstra, Devonport implemented Dubber Moments to pinpoint key moments of complaint with the AI within council calls and so proactively identify service issues within the community. Across 2000 calls, Dubber Moments flagged complaints with 90% accuracy, and detected that overall 15% of the calls Devonport receives were complaints (a difficult percentage to detect and analyse manually). As a result, Dubber Moments helped Devonport unlock insights about the needs of the customer and improve service to their community. Dubber’s technology doesn’t just cover complaints, but detected and divided 85% of all Devonport customer calls into service categories. Conversation Intelligence can help councils run customer service teams more efficiently and influence allocation of resourcing and funding based on demand.

How AI ensures compliance

Regulations continue to grow more complex worldwide. Rules spanning consumer protection, data privacy, capital requirements, and more create a web difficult for businesses to consistently keep up with. Yet, failure to comply can result in major fines or reputational damage. Unfortunately, this is a problem across the board. A 2023 survey found just 53% of UK executives feel ‘very prepared’ for data privacy compliance, demonstrating the evolving complexity confronting businesses.

Much of this is down to human error. Humans can easily misinterpret vague legal terminology or overlook important details buried in lengthy regulations. Fatigue, distractions, and complexity can result in compliance personnel missing key information or failing to properly apply policies. This opens businesses up to regulatory and reputational risk.

Conversation AI can be the solution to these human limitations. Advanced natural language processing enables AI to listen in and cross reference voice data with the various stipulations of intricate regulations, legislation, and policies in real-time. This technology helps overburdened human compliance functions by giving a far better oversight of what happens in the conversations in their organisation, and advice on how to correctly handle customer information based on the situation, scaling compliance knowledge across the organisation.

Redefining the way we work

Beyond these features, Conversation AI can unlock a universe of potential applications in business. The technology can optimise customer service by understanding conversations and helping businesses proactively address customer needs. It can analyse product performance based on customer feedback and recommend improvements, generate insights from customer data to inform marketing campaigns, credit risk models, and more. Previously, advanced AI systems required massive budgets to build in-house. Now, Conversation Intelligence is accessible as an ‘out-the-box’ software solution, allowing businesses of all sizes to benefit and level the playing field.

Conversation AI represents a significant opportunity for workforces globally. This technology has been decades in the making, fuelled by continual advances in deep learning and natural language processing. While unguided Large Language Models may produce inconsistent or unvetted outputs, this tech takes a different approach to help avoid human error in managing exponentially complex conversation data. Conversation AI is set to change the ways we work, from writing our meeting actions for us to detecting key moments in conversations, allowing us to work more productively and more informed than ever before.

Iain-McCowan-Dubber-Director-of-AI-1
Director of AI at Dubber | + posts

Dr Iain McCowan is Director of AI at Dubber. Former co-founder and CTO of Notiv, Iain leads AI at Dubber with over 20 years of experience in applied machine learning, speech and audio signal processing, research, project management and commercialisation. He holds a Ph.D. from the Queensland University of Technology.