Skip to content

Generative AI for the Service World

DMG defines Generative AI as “a type of artificial intelligence that leverages deep-learning algorithms to produce new content (e.g., text, images, computer code, workflows, models, audio/music, and more). These models use neural networks to find the patterns and structure in the data and apply it to generate original content which is intended to appear as if it has been created by a human being.” Generative AI is ideal for service organizations due to its ability to create content in response to inquiries. Contact center and customer service technology vendors are applying generative AI to enhance their solutions, speed up implementation, and improve the accuracy of their findings. Most of these solutions are new, having been released during the past few months, but due to the inherent capabilities of these applications, they are maturing quickly. 

Training Data Is Key

The underlying training data leveraged by the solutions is essential for the application’s success. Generative AI technologies utilized in contact center applications are primarily being trained using large language models (LLMs), which can contain generic training data (public LLMs), be more specific to an industry or company (private LLMs), or be a combination of both. Initially, public LLMs were leveraged almost exclusively, including various iterations of OpenAI’s Generative Pre-trained Transformers (GPT-3, GPT-3.5, GPT-4), Google’s PaLM, Anthropic’s Claude, Cohere, Hugging Face, and others. Today, a growing number of contact center/customer service software vendors are building customized private models trained on contact center data because their clients are showing a strong preference for domain-specific and/or verticalized LLMs. Vendors are also: fine-tuning public models with their own data to improve accuracy; supporting the use of multiple LLMs simultaneously; or leveraging different ones based on use case. In addition, some enterprises are beginning to create proprietary LLMs, enabling them to have complete control of the data leveraged by generative AI technologies used in their organizations. Since generative AI creates its original content based on the training data provided, it’s imperative that any LLM used, public or private, contains targeted, tagged, and curated information appropriate for the task at hand.

Applications for Contact Centers/Customer Service

Generative AI has caught the attention and imaginations of vendors and enterprises, driving investments in a large and growing variety of applications for contact centers (sales, marketing, service, collections, technical support, etc.) and customer service organizations. Some of the more common applications for generative AI are:

  • Self-service solutions – utilized in intelligent virtual agents (IVAs) and bots for system testing, identifying and building intents, composing conversation responses, sentiment detection, and more
  • Omnichannel routing – enable behavioral, preference, data-driven, and propensity-based routing
  • Training – facilitate automated skills-based, “always-on”/real-time, and closed-loop training and coaching
  • Workforce management – identify and apply the best-fit algorithms for forecasting and scheduling; optimize scheduling of concurrent omnichannel interactions; improve the effectiveness of intraday management; enhance long-term planning; identify call flow arrival patterns, and more
  • Interaction analytics – enhance the identification of customer intents, improve accuracy of call transcripts, identify underlying reasons customers reach out to an organization, enhance sentiment/emotion analysis, pinpoint operational bottlenecks, improve redaction, create post-interaction summaries, and more
  • Quality management (QM)/automated QM – identify performance issues, improve compliance, evaluate agent performance, identify friction in the customer journey, recognize outstanding performance, identify at-risk agents, and more
  • Customer relationship management (CRM) – identify customer patterns, personalize offers, identify and act on customer preferences, automate case management, provide proprietary knowledge, perform case-based routing, and more
  • Real-time guidance (RTG) – provide intent-based guidance and next-best- actions, deliver context-based information/knowledge articles/scripts, improve automated responses, perform case management, and more 
  • Knowledge management – source content, generate frequently asked questions and other responses, identify content gaps and create knowledge assets, summarize enterprise-wide knowledge assets into usable real-time content, provide conversational semantic search, and more

Final Thoughts
Generative AI has already proven to be highly valuable for enterprises, particularly for contact centers and service organizations, due to its ability to generate content to address all types of situations. However, these solutions are dependent upon their LLMs, and companies need to pay as much attention to the data repositories that are used as the source of content for these solutions as they do to the solutions that are using them. To gain an in-depth understanding of how to apply Generative AI to your operating environment, as well as which applications are ready for use, please see DMG’s newest report, Generative AI: A New Paradigm for Contact Centers and Customer Service.