Generative AI Conversational design
I design intelligent, inclusive and trustworthy Generative AI assistants. Drawing on user experience best practice and close technical collaboration, I craft assistants rooted in real value through meaningful interactions.
My core principles for designing AI conversational flows
Human-centred intelligence
I distinguish between “natural” and “human-like” conversation design. An assistant should feel intuitive — not imitate people. This avoids uncanny experiences and builds trust.
Inclusive by design
Accessibility is rooted into every decision. I design with empathy, considering the needs of users with cognitive, visual or speech impairments.
Fail-safe flows
Assistants should never dead-end. Every path is built with clear error handling, recovery options and graceful exits, so users always feel guided, not stuck.
My process for designing conversational flows
Define Scope & Intent
Outline the topic, the assistant’s purpose and the goals users arrive with.
Script & Sentiment Mapping
Map dialogue, tone and actions to match user expectations and emotional context.
Collaborative Iteration
Work closely with developers and solution architects, diverse feedback makes generative flows more resilient.
Build, Tune & Test
Simulate and test conversations against real use cases, tuning intents and responses.
Validation
I validate logic and tone through regular playback sessions turning feedback into fast improvements.