Unified conversational support experience
Unified conversational support experience
Unified conversational support experience
Created Intel’s first unified AI assistant experience, replacing fragmented support pathways with a guided conversational interface that improves task discovery and reduces user dropout.
Enterprise-ready conversational architecture
Enterprise-ready conversational architecture
Enterprise-ready conversational architecture
Designed a scalable conversational framework with reusable chat patterns, escalation logic and message structures enabling consistent AI experiences across Intel.com support channels.
Frictionless task completion and clearer guidance
Frictionless task completion and clearer guidance
Frictionless task completion and clearer guidance
Streamlined complex troubleshooting flows into clear, step-by-step conversational paths, reducing confusion and improving task completion across support scenarios.
Overview
Ask Intel is the company’s first enterprise-wide virtual assistant designed to help users find support, troubleshoot issues and access Intel resources across multiple channels. I led the end-to-end design of this new AI-powered experience defining the conversational model, building the chat UI and creating the full system of components and patterns.
This was a ground-up effort: no prior assistant, no existing framework and no shared model for conversational behavior. I established the foundation that Intel’s conversational experiences will scale on for years to come.
My Role
Lead Product Designer on Digital Experiences Team
As Lead User Experience Designer for Intel.com, I owned the design for the new Ask Intel conversational AI experience. We made getting technical support at Intel incredibly simple. Before, customers faced lengthy documentation searches and wait times for live support. The digital experiences team and I changed that. We focused on our users. We found their most common questions. Then we designed, tested and built a solution. Our work was always guided by accessibility and privacy transparency. It was built for a global audience with diverse technical needs. The final product is an intelligent virtual assistant. It instantly connects customers to answers while maintaining a clear path to human support when needed. It also provides a blueprint for conversational AI experiences at Intel.
The Problem
Have you ever needed a quick answer to a technical question, but finding it meant digging through dense documentation or waiting on hold for support? That's what was happening at Intel. Customers had straightforward questions about processors, compatibility and troubleshooting, but getting answers was time-consuming and frustrating. We had extensive documentation and live support, but no intelligent middle ground that could instantly help with common issues.
Fragmented support pathways
Users bounced between product pages, documentation, forums and contact forms trying to find answers to common technical questions.
Long wait times for simple queries
Live support agents spent significant time answering repetitive questions that could be automated, while customers waited unnecessarily.
Poor mobile support experience
Technical documentation wasn't optimized for mobile, yet many users were trying to troubleshoot issues on their phones.
Missed self-service opportunity
There was no intelligent system to guide users to the right information based on their specific needs and context.
What actually happened
We started by analyzing support ticket data and talking to customer support teams. This helped us understand the most common questions and pain points. By identifying patterns in user inquiries, we could design conversation flows that felt natural and helpful. We weren't trying to replace human support, we were creating an intelligent first line of assistance that could instantly help with common issues while seamlessly escalating complex cases to live agents.
The Solution
We structured interactions around topic categories with clear quick-reply buttons to guide users efficiently. We designed clear escalation paths so users could connect with live agents when needed and we surfaced data usage transparency upfront with a clear disclaimer before engagement.
Wireframing and Design System Integration
We started with low-fidelity wireframes to map out conversation flows and button placement, testing the structure to identify where users might get stuck. Then we built every component using our Atomic Design System, ensuring visual consistency with Intel.com and seamless responsive experiences from desktop to mobile.
Enhanced User Experiences
Pre-structured topic buttons help users navigate quickly without needing to phrase questions perfectly, the conversation unfolds logically with progressive disclosure, the interface works seamlessly across devices for mobile troubleshooting and technical information is presented in plain language while maintaining accuracy.
Implementation
We reviewed support tickets to identify common inquiries, collaborated with customer support teams and product managers on backend capabilities, documented conversation flows and escalation triggers and ensured button-based navigation with keyboard and screen reader compatibility meeting WCAG 2.1 AA standards.
Results
The Ask Intel virtual assistant launched successfully across Intel’s support ecosystem and became a central entry point for customers seeking help. Key outcomes include:
Users found answers faster and with less friction, avoiding the need to navigate multiple disconnected support systems.
Support agents spent less time addressing routine, repetitive questions, allowing them to focus on high-complexity issues.
The assistant created a unified support entry point, strengthening consistency across business units.
Adoption grew organically because the experience felt easier than traditional search.
Content owners gained a clearer governance model for maintaining and improving support knowledge.
The assistant established the foundation for future AI-driven support experiences within Intel’s ecosystem.
What I learned
Conversational AI is as much about systems design as it is about UI. Designing an assistant requires aligning taxonomy, content quality, NLP capabilities, localization and user expectations, not just building a chat window.
Enterprise users expect clarity, predictability and trust. Conversational design must be transparent, concise and free of ambiguity to work effectively inside large organizations.
Early testing is invaluable. Accessibility, non-native language use and failure states surfaced issues no static prototype could reveal.
Design system integration is the backbone of scalability. By contributing conversational components to the Atomic Design System, future teams could adopt and expand the assistant without reinventing the foundation.
If you're building AI-enabled experiences, modern design systems or complex UX at scale, I can help.
Overview
Ask Intel is the company’s first enterprise-wide virtual assistant designed to help users find support, troubleshoot issues and access Intel resources across multiple channels. I led the end-to-end design of this new AI-powered experience defining the conversational model, building the chat UI and creating the full system of components and patterns.
This was a ground-up effort: no prior assistant, no existing framework and no shared model for conversational behavior. I established the foundation that Intel’s conversational experiences will scale on for years to come.
My Role
Lead Product Designer on Digital Experiences Team
As Lead User Experience Designer for Intel.com, I owned the design for the new Ask Intel conversational AI experience. We made getting technical support at Intel incredibly simple. Before, customers faced lengthy documentation searches and wait times for live support. The digital experiences team and I changed that. We focused on our users. We found their most common questions. Then we designed, tested and built a solution. Our work was always guided by accessibility and privacy transparency. It was built for a global audience with diverse technical needs. The final product is an intelligent virtual assistant. It instantly connects customers to answers while maintaining a clear path to human support when needed. It also provides a blueprint for conversational AI experiences at Intel.
The Problem
Have you ever needed a quick answer to a technical question, but finding it meant digging through dense documentation or waiting on hold for support? That's what was happening at Intel. Customers had straightforward questions about processors, compatibility and troubleshooting, but getting answers was time-consuming and frustrating. We had extensive documentation and live support, but no intelligent middle ground that could instantly help with common issues.
Fragmented support pathways
Users bounced between product pages, documentation, forums and contact forms trying to find answers to common technical questions.
Long wait times for simple queries
Live support agents spent significant time answering repetitive questions that could be automated, while customers waited unnecessarily.
Poor mobile support experience
Technical documentation wasn't optimized for mobile, yet many users were trying to troubleshoot issues on their phones.
Missed self-service opportunity
There was no intelligent system to guide users to the right information based on their specific needs and context.
What actually happened
We started by analyzing support ticket data and talking to customer support teams. This helped us understand the most common questions and pain points. By identifying patterns in user inquiries, we could design conversation flows that felt natural and helpful. We weren't trying to replace human support, we were creating an intelligent first line of assistance that could instantly help with common issues while seamlessly escalating complex cases to live agents.
The Solution
We structured interactions around topic categories with clear quick-reply buttons to guide users efficiently. We designed clear escalation paths so users could connect with live agents when needed and we surfaced data usage transparency upfront with a clear disclaimer before engagement.
Wireframing and Design System Integration
We started with low-fidelity wireframes to map out conversation flows and button placement, testing the structure to identify where users might get stuck. Then we built every component using our Atomic Design System, ensuring visual consistency with Intel.com and seamless responsive experiences from desktop to mobile.
Enhanced User Experiences
Pre-structured topic buttons help users navigate quickly without needing to phrase questions perfectly, the conversation unfolds logically with progressive disclosure, the interface works seamlessly across devices for mobile troubleshooting and technical information is presented in plain language while maintaining accuracy.
Implementation
We reviewed support tickets to identify common inquiries, collaborated with customer support teams and product managers on backend capabilities, documented conversation flows and escalation triggers and ensured button-based navigation with keyboard and screen reader compatibility meeting WCAG 2.1 AA standards.
Results
The Ask Intel virtual assistant launched successfully across Intel’s support ecosystem and became a central entry point for customers seeking help. Key outcomes include:
Users found answers faster and with less friction, avoiding the need to navigate multiple disconnected support systems.
Support agents spent less time addressing routine, repetitive questions, allowing them to focus on high-complexity issues.
The assistant created a unified support entry point, strengthening consistency across business units.
Adoption grew organically because the experience felt easier than traditional search.
Content owners gained a clearer governance model for maintaining and improving support knowledge.
The assistant established the foundation for future AI-driven support experiences within Intel’s ecosystem.
What I learned
Conversational AI is as much about systems design as it is about UI. Designing an assistant requires aligning taxonomy, content quality, NLP capabilities, localization and user expectations, not just building a chat window.
Enterprise users expect clarity, predictability and trust. Conversational design must be transparent, concise and free of ambiguity to work effectively inside large organizations.
Early testing is invaluable. Accessibility, non-native language use and failure states surfaced issues no static prototype could reveal.
Design system integration is the backbone of scalability. By contributing conversational components to the Atomic Design System, future teams could adopt and expand the assistant without reinventing the foundation.
Overview
Ask Intel is the company’s first enterprise-wide virtual assistant designed to help users find support, troubleshoot issues and access Intel resources across multiple channels. I led the end-to-end design of this new AI-powered experience defining the conversational model, building the chat UI and creating the full system of components and patterns.
This was a ground-up effort: no prior assistant, no existing framework and no shared model for conversational behavior. I established the foundation that Intel’s conversational experiences will scale on for years to come.
My Role
Lead Product Designer on Digital Experiences Team
As Lead User Experience Designer for Intel.com, I owned the design for the new Ask Intel conversational AI experience. We made getting technical support at Intel incredibly simple. Before, customers faced lengthy documentation searches and wait times for live support. The digital experiences team and I changed that. We focused on our users. We found their most common questions. Then we designed, tested and built a solution. Our work was always guided by accessibility and privacy transparency. It was built for a global audience with diverse technical needs. The final product is an intelligent virtual assistant. It instantly connects customers to answers while maintaining a clear path to human support when needed. It also provides a blueprint for conversational AI experiences at Intel.
The Problem
Have you ever needed a quick answer to a technical question, but finding it meant digging through dense documentation or waiting on hold for support? That's what was happening at Intel. Customers had straightforward questions about processors, compatibility and troubleshooting, but getting answers was time-consuming and frustrating. We had extensive documentation and live support, but no intelligent middle ground that could instantly help with common issues.
Fragmented support pathways
Users bounced between product pages, documentation, forums and contact forms trying to find answers to common technical questions.
Long wait times for simple queries
Live support agents spent significant time answering repetitive questions that could be automated, while customers waited unnecessarily.
Poor mobile support experience
Technical documentation wasn't optimized for mobile, yet many users were trying to troubleshoot issues on their phones.
Missed self-service opportunity
There was no intelligent system to guide users to the right information based on their specific needs and context.
What actually happened
We started by analyzing support ticket data and talking to customer support teams. This helped us understand the most common questions and pain points. By identifying patterns in user inquiries, we could design conversation flows that felt natural and helpful. We weren't trying to replace human support, we were creating an intelligent first line of assistance that could instantly help with common issues while seamlessly escalating complex cases to live agents.
The Solution
We structured interactions around topic categories with clear quick-reply buttons to guide users efficiently. We designed clear escalation paths so users could connect with live agents when needed and we surfaced data usage transparency upfront with a clear disclaimer before engagement.
Wireframing and Design System Integration
We started with low-fidelity wireframes to map out conversation flows and button placement, testing the structure to identify where users might get stuck. Then we built every component using our Atomic Design System, ensuring visual consistency with Intel.com and seamless responsive experiences from desktop to mobile.
Enhanced User Experiences
Pre-structured topic buttons help users navigate quickly without needing to phrase questions perfectly, the conversation unfolds logically with progressive disclosure, the interface works seamlessly across devices for mobile troubleshooting and technical information is presented in plain language while maintaining accuracy.
Implementation
We reviewed support tickets to identify common inquiries, collaborated with customer support teams and product managers on backend capabilities, documented conversation flows and escalation triggers and ensured button-based navigation with keyboard and screen reader compatibility meeting WCAG 2.1 AA standards.
Results
The Ask Intel virtual assistant launched successfully across Intel’s support ecosystem and became a central entry point for customers seeking help. Key outcomes include:
Users found answers faster and with less friction, avoiding the need to navigate multiple disconnected support systems.
Support agents spent less time addressing routine, repetitive questions, allowing them to focus on high-complexity issues.
The assistant created a unified support entry point, strengthening consistency across business units.
Adoption grew organically because the experience felt easier than traditional search.
Content owners gained a clearer governance model for maintaining and improving support knowledge.
The assistant established the foundation for future AI-driven support experiences within Intel’s ecosystem.
What I learned
Conversational AI is as much about systems design as it is about UI. Designing an assistant requires aligning taxonomy, content quality, NLP capabilities, localization and user expectations, not just building a chat window.
Enterprise users expect clarity, predictability and trust. Conversational design must be transparent, concise and free of ambiguity to work effectively inside large organizations.
Early testing is invaluable. Accessibility, non-native language use and failure states surfaced issues no static prototype could reveal.
Design system integration is the backbone of scalability. By contributing conversational components to the Atomic Design System, future teams could adopt and expand the assistant without reinventing the foundation.
Overview
Ask Intel is the company’s first enterprise-wide virtual assistant designed to help users find support, troubleshoot issues and access Intel resources across multiple channels. I led the end-to-end design of this new AI-powered experience defining the conversational model, building the chat UI and creating the full system of components and patterns.
This was a ground-up effort: no prior assistant, no existing framework and no shared model for conversational behavior. I established the foundation that Intel’s conversational experiences will scale on for years to come.
My Role
Lead Product Designer on Digital Experiences Team
As Lead User Experience Designer for Intel.com, I owned the design for the new Ask Intel conversational AI experience. We made getting technical support at Intel incredibly simple. Before, customers faced lengthy documentation searches and wait times for live support. The digital experiences team and I changed that. We focused on our users. We found their most common questions. Then we designed, tested and built a solution. Our work was always guided by accessibility and privacy transparency. It was built for a global audience with diverse technical needs. The final product is an intelligent virtual assistant. It instantly connects customers to answers while maintaining a clear path to human support when needed. It also provides a blueprint for conversational AI experiences at Intel.
The Problem
Have you ever needed a quick answer to a technical question, but finding it meant digging through dense documentation or waiting on hold for support? That's what was happening at Intel. Customers had straightforward questions about processors, compatibility and troubleshooting, but getting answers was time-consuming and frustrating. We had extensive documentation and live support, but no intelligent middle ground that could instantly help with common issues.
Fragmented support pathways
Users bounced between product pages, documentation, forums and contact forms trying to find answers to common technical questions.
Long wait times for simple queries
Live support agents spent significant time answering repetitive questions that could be automated, while customers waited unnecessarily.
Poor mobile support experience
Technical documentation wasn't optimized for mobile, yet many users were trying to troubleshoot issues on their phones.
Missed self-service opportunity
There was no intelligent system to guide users to the right information based on their specific needs and context.
What actually happened
We started by analyzing support ticket data and talking to customer support teams. This helped us understand the most common questions and pain points. By identifying patterns in user inquiries, we could design conversation flows that felt natural and helpful. We weren't trying to replace human support, we were creating an intelligent first line of assistance that could instantly help with common issues while seamlessly escalating complex cases to live agents.
The Solution
We structured interactions around topic categories with clear quick-reply buttons to guide users efficiently. We designed clear escalation paths so users could connect with live agents when needed and we surfaced data usage transparency upfront with a clear disclaimer before engagement.
Wireframing and Design System Integration
We started with low-fidelity wireframes to map out conversation flows and button placement, testing the structure to identify where users might get stuck. Then we built every component using our Atomic Design System, ensuring visual consistency with Intel.com and seamless responsive experiences from desktop to mobile.
Enhanced User Experiences
Pre-structured topic buttons help users navigate quickly without needing to phrase questions perfectly, the conversation unfolds logically with progressive disclosure, the interface works seamlessly across devices for mobile troubleshooting and technical information is presented in plain language while maintaining accuracy.
Implementation
We reviewed support tickets to identify common inquiries, collaborated with customer support teams and product managers on backend capabilities, documented conversation flows and escalation triggers and ensured button-based navigation with keyboard and screen reader compatibility meeting WCAG 2.1 AA standards.
Results
The Ask Intel virtual assistant launched successfully across Intel’s support ecosystem and became a central entry point for customers seeking help. Key outcomes include:
Users found answers faster and with less friction, avoiding the need to navigate multiple disconnected support systems.
Support agents spent less time addressing routine, repetitive questions, allowing them to focus on high-complexity issues.
The assistant created a unified support entry point, strengthening consistency across business units.
Adoption grew organically because the experience felt easier than traditional search.
Content owners gained a clearer governance model for maintaining and improving support knowledge.
The assistant established the foundation for future AI-driven support experiences within Intel’s ecosystem.
What I learned
Conversational AI is as much about systems design as it is about UI. Designing an assistant requires aligning taxonomy, content quality, NLP capabilities, localization and user expectations, not just building a chat window.
Enterprise users expect clarity, predictability and trust. Conversational design must be transparent, concise and free of ambiguity to work effectively inside large organizations.
Early testing is invaluable. Accessibility, non-native language use and failure states surfaced issues no static prototype could reveal.
Design system integration is the backbone of scalability. By contributing conversational components to the Atomic Design System, future teams could adopt and expand the assistant without reinventing the foundation.
Overview
Ask Intel is the company’s first enterprise-wide virtual assistant designed to help users find support, troubleshoot issues and access Intel resources across multiple channels. I led the end-to-end design of this new AI-powered experience defining the conversational model, building the chat UI and creating the full system of components and patterns.
This was a ground-up effort: no prior assistant, no existing framework and no shared model for conversational behavior. I established the foundation that Intel’s conversational experiences will scale on for years to come.
My Role
Lead Product Designer on Digital Experiences Team
As Lead User Experience Designer for Intel.com, I owned the design for the new Ask Intel conversational AI experience. We made getting technical support at Intel incredibly simple. Before, customers faced lengthy documentation searches and wait times for live support. The digital experiences team and I changed that. We focused on our users. We found their most common questions. Then we designed, tested and built a solution. Our work was always guided by accessibility and privacy transparency. It was built for a global audience with diverse technical needs. The final product is an intelligent virtual assistant. It instantly connects customers to answers while maintaining a clear path to human support when needed. It also provides a blueprint for conversational AI experiences at Intel.
The Problem
Have you ever needed a quick answer to a technical question, but finding it meant digging through dense documentation or waiting on hold for support? That's what was happening at Intel. Customers had straightforward questions about processors, compatibility and troubleshooting, but getting answers was time-consuming and frustrating. We had extensive documentation and live support, but no intelligent middle ground that could instantly help with common issues.
Fragmented support pathways
Users bounced between product pages, documentation, forums and contact forms trying to find answers to common technical questions.
Long wait times for simple queries
Live support agents spent significant time answering repetitive questions that could be automated, while customers waited unnecessarily.
Poor mobile support experience
Technical documentation wasn't optimized for mobile, yet many users were trying to troubleshoot issues on their phones.
Missed self-service opportunity
There was no intelligent system to guide users to the right information based on their specific needs and context.
What actually happened
We started by analyzing support ticket data and talking to customer support teams. This helped us understand the most common questions and pain points. By identifying patterns in user inquiries, we could design conversation flows that felt natural and helpful. We weren't trying to replace human support, we were creating an intelligent first line of assistance that could instantly help with common issues while seamlessly escalating complex cases to live agents.
The Solution
We structured interactions around topic categories with clear quick-reply buttons to guide users efficiently. We designed clear escalation paths so users could connect with live agents when needed and we surfaced data usage transparency upfront with a clear disclaimer before engagement.
Wireframing and Design System Integration
We started with low-fidelity wireframes to map out conversation flows and button placement, testing the structure to identify where users might get stuck. Then we built every component using our Atomic Design System, ensuring visual consistency with Intel.com and seamless responsive experiences from desktop to mobile.
Enhanced User Experiences
Pre-structured topic buttons help users navigate quickly without needing to phrase questions perfectly, the conversation unfolds logically with progressive disclosure, the interface works seamlessly across devices for mobile troubleshooting and technical information is presented in plain language while maintaining accuracy.
Implementation
We reviewed support tickets to identify common inquiries, collaborated with customer support teams and product managers on backend capabilities, documented conversation flows and escalation triggers and ensured button-based navigation with keyboard and screen reader compatibility meeting WCAG 2.1 AA standards.
Results
The Ask Intel virtual assistant launched successfully across Intel’s support ecosystem and became a central entry point for customers seeking help. Key outcomes include:
Users found answers faster and with less friction, avoiding the need to navigate multiple disconnected support systems.
Support agents spent less time addressing routine, repetitive questions, allowing them to focus on high-complexity issues.
The assistant created a unified support entry point, strengthening consistency across business units.
Adoption grew organically because the experience felt easier than traditional search.
Content owners gained a clearer governance model for maintaining and improving support knowledge.
The assistant established the foundation for future AI-driven support experiences within Intel’s ecosystem.
What I learned
Conversational AI is as much about systems design as it is about UI. Designing an assistant requires aligning taxonomy, content quality, NLP capabilities, localization and user expectations, not just building a chat window.
Enterprise users expect clarity, predictability and trust. Conversational design must be transparent, concise and free of ambiguity to work effectively inside large organizations.
Early testing is invaluable. Accessibility, non-native language use and failure states surfaced issues no static prototype could reveal.
Design system integration is the backbone of scalability. By contributing conversational components to the Atomic Design System, future teams could adopt and expand the assistant without reinventing the foundation.