Growing a Conversation Design Practice

In this episode of Unprompted, Hans van Dam speaks with Grace Hughes, Conversation Design Director at Accenture Song, about growing a conversation design practice within enterprise environments. The discussion explores how the role of the conversation designer is evolving, how organizations build design capabilities at scale, and what it takes to maintain quality and trust in AI systems.

The Evolution of Conversation Design

Conversation design has evolved significantly over the past decade. Early work focused heavily on scripting dialogues and designing structured flows. As conversational systems become more complex and more generative, the role of the designer has shifted.

Grace explains how conversation designers today influence broader product behavior. Rather than writing every line of dialogue, designers increasingly define principles, interaction patterns, and guardrails that guide how AI systems behave across many unpredictable interactions.

Designing in the LLM Era

Large language models have introduced new flexibility but also new responsibilities. Designers are no longer simply controlling conversations. Instead, they must shape systems that can generate responses dynamically while still remaining aligned with user expectations.

This requires designing boundaries, defining acceptable behavior, and ensuring that systems remain understandable and trustworthy even when they operate in non deterministic ways.

Building a Design Practice

Growing a conversation design practice inside a large organization requires more than individual expertise. It requires structure.

Grace discusses how organizations establish Centers of Excellence to support shared practices, design standards, and governance. These structures help teams scale expertise across multiple products and departments while maintaining consistent quality.

Building a practice also requires creating alignment between design, engineering, product management, and research.

Maintaining Quality at Scale

As conversational systems expand across platforms and use cases, maintaining consistency becomes a major challenge.

The conversation highlights the importance of shared frameworks, documentation, and collaboration across teams. These mechanisms allow organizations to maintain quality while enabling teams to move quickly and experiment responsibly.

Quality in conversational AI is not only about language. It is about system behavior, transparency, and user confidence.

Designing for Trust

Trust plays a central role in conversational experiences. Users must understand what the system can do and where its limitations lie.

Grace explains that designing for trust involves transparency, careful language choices, and responsible interaction patterns. Designers must help users feel confident without overstating the capabilities of AI systems.

As conversational interfaces become more embedded in everyday products, trust becomes one of the most important design goals.

Conversation Design as Systems Thinking

The episode highlights how inversation design has expanded beyond dialogue writing.

Today’s practitioners must understand product strategy, system architecture, governance, and collaboration across disciplines. Conversation design increasingly sits at the intersection of design, technology, and organizational structure.

Grace emphasizes that building successful conversational systems requires strong collaboration and continuous learning.

Meet the Speakers

Grace Hughes


Conversation Design Director, Accenture Song

Grace Hughes leads conversation design strategy and execution at Accenture Song, working on large scale conversational and content design initiatives. Her work spans digital identity frameworks, multimodal AI assistants, and enterprise conversational platforms.

Grace is also an educator and mentor in the field. She teaches at TU Dublin and IADT and contributes to research and industry reports on conversational AI and value networks. Her work focuses on scaling human centered design practices across complex organizations.


Hans van Dam

CEO and Co founder, Conversation Design Institute

Hans van Dam co founded the Conversation Design Institute to professionalize the role of conversation design within large enterprises. CDI provides training, certification, and advisory services that help organizations build and scale conversational AI capabilities.

Hans focuses on connecting design practice with platform strategy and organizational governance.

Watch the Webinar

This blog highlights key ideas from the conversation, but the full discussion offers deeper insights and practical examples from Grace’s experience building conversation design practices inside large organizations.

Watch the complete episode of Unprompted below.