Nuance Mix
Designing a scalable platform for conversational AI
Nuance Mix is a platform for creating, training, and governing conversational AI across voice and digital channels.
My work focused on shaping the underlying systems from project creation and dialog design to NLU, governance, and internal tooling so the platform could grow in capability without losing clarity.
Year
03.20 - 08.24
Impact
Timeline
3.6 years
Nuance Mix is a platform for designing, training, and deploying conversational AI across voice and digital channels. I worked across multiple foundational areas of the product including project creation, NLU, dialog design, design systems, governance, and internal customer success tools with a focus on ensuring that a powerful, technical system remained clear, consistent, and usable at scale. Rather than optimizing isolated features, my work centered on shaping how the platform fits together.
Across project creation, NLU, dialog authoring, governance, design systems, and internal support tools, my work focused on making a technically complex conversational-AI platform easier to understand and operate. I shaped shared patterns and lifecycle clarity so teams could move from experimentation to production with more confidence.
Mix project creation
Opportunity
Conversational AI platforms are inherently complex.
They sit at the intersection of:
language models
intent classification
dialog orchestration
enterprise governance
and real-world operational constraints
As Mix expanded, new capabilities were added quickly.
The risk wasn’t lack of power, it was fragmentation.
The challenge was to allow the platform to grow without losing clarity or coherence.
NLU
Design Intent
I approached Mix with a simple philosophy:
Make complexity legible
Design for scale, not demos
Treat governance as a core experience
Ensure creators understand the system they’re shaping
The goal was not to hide complexity, but to organize it thoughtfully.
Dialog
What I Owned
Project Creation & Structure
I designed how users create and structure projects, the starting point for everything that follows.
Focus areas:
Clear project setup and configuration
Understandable defaults
Predictable structure that supports growth over time
This ensured teams could move from experimentation to production without rework.
NLU & Intent Design
I worked on the experiences around defining, training, and validating intents and entities.
Key considerations:
Making abstract language concepts concrete
Helping users understand model behaviour
Reducing errors during training and testing
The goal was to give creators confidence in how the system interprets language.
Dialog Design
I contributed to dialog authoring i.e. how conversations are structured, branched, and maintained.
Design priorities:
Clear mental models for dialog flow
Scalable patterns for complex conversations
Editing experiences that remain understandable as logic grows
Dialog design needed to feel deliberate, not brittle.
Design Systems & Platform Consistency
Alongside feature work, I helped establish shared design patterns and system guidelines across Mix.
This included:
Visual and interaction consistency
Reusable components
Predictable behaviours across surfaces
Consistency was critical for a platform used daily by expert users.
Governance & Enterprise Controls
I worked on governance-related experiences to ensure Mix could operate in regulated, enterprise environments.
This included:
Permissions and access control patterns
Clear system states
Predictable publishing and lifecycle behaviours
Governance was treated as a first-class design problem, not a constraint.
Internal Customer Success Tools (Frontend Development)
I also built frontend tooling used by internal customer success teams.
This work focused on:
Improving visibility into customer configurations
Reducing manual support workflows
Enabling faster diagnosis and resolution
It grounded my design decisions in real operational needs.
Impact
Across the platform areas I worked on, the aim was to make advanced conversational-AI workflows more understandable and dependable:
Reduced errors in NLU and dialog configuration
Faster onboarding for new users
Stronger consistency across the platform
Better operational efficiency for internal support teams
My contribution was to help Mix feel like a coherent platform rather than a collection of advanced tools.cohesive platform.
Collaboration & Leadership:
This work required close collaboration with:
PMs defining platform direction
Engineers building NLU and dialog infrastructure
Customer success teams supporting enterprise deployments
Design partners across Nuance
I frequently:
aligned teams on shared mental models
reviewed designs for cross-surface consistency
identified systemic gaps early
balanced user needs with enterprise constraints
Much of the value came from identifying systemic complexity early, before it became a customer problem.preventing complexity before it surfaced.
Reflection
Designing Nuance Mix wasn’t about crafting a single standout feature.
It was about shaping a system that could support serious, real-world conversational AI work reliably and at scale.
The best outcome of this project was quiet:
fewer questions
fewer errors
smoother handoffs
and a platform that felt intentional
