Project Overview
The restaurant industry is rapidly evolving, with automation playing a key role in addressing staffing shortages and streamlining operations. Virnika is developing an AI-powered voice ordering system for small and medium-sized restaurants, providing a white-label solution to enhance accessibility and reduce the burden on staff.
Tools: Figma, FigJam, Photoshop, Illustrator, Apple Motion, Pen & Paper
Timeline: Three-week sprint
My Role
Voice UI & Interaction Design
Competitive Research & Strategy
Prototyping & AI Visualization
My Responsibilities:
Designed intuitive voice interactions and user flows for AI-driven restaurant ordering.
Conducted market analysis to benchmark industry trends and identify opportunities for AI-powered automation.
Led the development of interactive prototypes and designed voice interface visualizations to enhance user experience.
Understanding the Problem
Why Voice AI for Restaurants?
The rise of automation in food service is driven by:
Staffing shortages and high turnover in restaurants.
Growing customer demand for faster, more personalized ordering experiences.
Opportunities for improved accessibility for visually impaired or multitasking users.
Research Goals
We focused on three primary questions:
1️⃣ What challenges do restaurant owners and customers face with online ordering?
2️⃣ How can AI-driven voice interactions reduce friction in the ordering process?
3️⃣ What existing solutions exist, and where can Virnika provide unique value?
User Research & Insights
User Interviews: Identifying Pain Points
We conducted five interviews via Zoom with restaurant staff and small business owners to uncover pain points with online ordering and automation. This qualitative research helped define common frustrations, feature expectations, and accessibility needs.
🗝️ Key Takeaways
Users dislike hidden fees and complex menus that slow down ordering.
Many customers want personalized recommendations based on past orders.
Business owners see potential in AI automation but need it to be accurate, efficient, and seamless.
How This Informed Design
✔ Prioritized menu transparency to eliminate confusion about pricing.
✔ Introduced smart recommendations based on past orders to streamline decision-making.
✔ Ensured voice AI feedback was clear and responsive, so users feel confident in their orders.
Market & Competitor Research
SWOT Analysis
Since few direct competitors exist for AI-powered restaurant ordering, we analyzed adjacent technologies that use automation in food service. This helped identify strengths, weaknesses, and opportunities for Virnika’s design.
Key Insights 🔦
Kea.ai
Strengths: AI-driven order processing, restaurant partnerships
Weaknesses: Low brand recognition, ongoing AI learning curve
SoundHound AI
Strengths: Strong AI capabilities, multilingual support
Weaknesses: High costs, limited restaurant adoption
Toast
Strengths: Industry-standard POS, strong third-party integrations
Weaknesses: Hardware-dependent, costly updates
Presto
Strengths: Scalable AI, employee satisfaction focus
Weaknesses: Customer resistance, accessibility concerns
Full SWOT Analysis:
How This Informed Design
✔ Focused on making AI voice interactions feel natural rather than robotic.
✔ Designed a low-friction onboarding process for restaurant owners.
✔ Ensured full compliance with accessibility standards (WCAG, ADA).
Synthesizing Research into Design
Affinity Mapping & User Journey Insights
To synthesize our research, we conducted affinity mapping to identify patterns in user behavior and pain points. This led to a user journey map outlining the typical restaurant customer’s ordering experience.
Example Affinity Map Insights 🔍
Key Findings from Affinity Mapping 🔑
Users experience decision fatigue when ordering from large, complex menus.
Customers expect real-time feedback from AI to confirm accuracy.
Restaurant owners prioritize efficiency over AI sophistication.
How This Informed Design
✔ Streamlined the ordering flow to reduce steps and minimize decision fatigue.
✔ Designed an AI-driven voice feedback loop to confirm orders in real-time.
✔ Prioritized simplicity and efficiency to ensure smooth adoption for restaurant owners.
User Insights: The Budget-Conscious Diner
Through user interviews and affinity mapping, we identified a key user type:
Adrian Brohdee – The Budget-Conscious Diner
I'm making alright money, but with the economy these days, I'm keeping a close eye on how much I spend when ordering food online.
Key Behaviors & Needs:
Prefers local restaurants and mobile apps for convenience.
Gets overwhelmed by extensive menus and dislikes hidden fees.
Values personalized recommendations based on past orders.
Expects efficient, human-like AI interactions for seamless ordering.
How This Informed Design
✔ Transparent pricing to eliminate fee confusion.
✔ AI-powered recommendations based on order history.
✔ Minimalist menu structure to reduce cognitive overload.
✔ Natural-sounding voice interactions to enhance trust & usability.
User Journey: From Consideration to Order Completion
Using user interviews, competitive analysis, and affinity mapping, we mapped the end-to-end ordering experience to identify pain points and opportunities for improvement.
Key Journey Stages:
Plan – Users decide where to order based on budget, past experiences, or recommendations.
Consider – They browse menus but may feel overwhelmed by too many options or unclear pricing.
Order – Users place an order, expecting accuracy, efficiency, and minimal friction.
Delivery – They anticipate real-time updates and smooth handoff of their order.
Enjoy – Users engage with their meal, noting whether the experience met expectations.
Feedback – They may leave reviews, reorder from favorites, or abandon platforms with poor experiences.
How This Informed Design
✔ Simplified menu structure to reduce decision fatigue.
✔ Transparent pricing & fee breakdowns to eliminate surprises.
✔ AI-driven order confirmations & recommendations to enhance confidence.
✔ Seamless voice interactions to streamline navigation & checkout.
UI Design Process: From Sketches to Wireframes
Before refining the voice assistant animation, we first focused on designing the overall UI for the ordering experience.
Design Studio Session
As a team, we conducted a collaborative sketching session to rapidly explore different interface layouts. Each of us created low-fidelity sketches, focusing on:
How the AI voice interaction would be visually represented.
The most intuitive menu & checkout flow for users.
Ensuring accessibility & ease of navigation.
Wireframe Development
After evaluating our initial concepts, we refined the design direction in wireframes. Some early ideas were reconsidered due to usability constraints, ultimately leading to a more structured, user-friendly layout.
Designing the AI Voice Interface
To create an intuitive AI voice experience, I explored various animation styles, drawing inspiration from Siri and other voice interfaces. Through iterative prototyping, user testing, and accessibility research, we refined the visual feedback system to enhance user confidence and engagement.
Iterative Animation Testing
Using Figma & Illustrator, I developed early animation concepts, later refining them in Apple Motion for usability testing. The goal was to design a clear, responsive interaction that effectively communicates active listening and processing.
To ensure the best user experience, I:
Researched best practices in AI voice UI animation (Siri, Google Assistant)
Developed interactive prototypes in Figma, Illustrator, and Apple Motion
Conducted usability tests to refine waveform responsiveness & accessibility
Final Two Design Options
Usability testing revealed two strong contenders:
Waveform encased in an orb – A fluid, spherical animation indicating active listening.
Horizontal line waveform – A minimalistic, real-time visualization of speech input.
To simulate real-world interactions, I synced these animations with recorded voice samples, demonstrating how users would experience AI feedback in real time.
Final Design Selection - Why the Waveform Won 🏆
Through usability testing and stakeholder feedback, the waveform animation was chosen for its:
Clear indication of active listening, increasing user confidence.
Dynamic responsiveness, making interactions feel natural.
Minimalist, intuitive UI, reducing cognitive load and improving accessibility.
After testing multiple design iterations, users preferred the waveform animation for its clarity, responsiveness, and ease of understanding. The final version was integrated into usability tests and stakeholder presentations to validate its effectiveness.
Final Results & Next Steps
Key Outcomes
✔ Reduced cognitive load by simplifying menu navigation.
✔ Improved AI accuracy with real-time voice feedback loops.
✔ Optimized accessibility through compliance with WCAG & ADA guidelines.
Next Steps
✔ Further usability testing to refine voice command structures.
✔ Optimizing AI responses to reduce errors and friction.
✔ Expanding accessibility features for a more inclusive user experience.
Final Prototype & Takeaways
With findings from research and usability testing, I refined the final prototype to showcase an intuitive, voice-first ordering experience that balanced clarity, efficiency, and accessibility.
The Final Design Introduced
✔ A simplified call flow interface, visually mapping out steps in the ordering experience for both users and restaurant staff.
✔ Voice interface visualizations to communicate system status, confirmations, and next steps.
✔ Accessibility-informed layouts and large touch targets for alternate navigation modes.
While the product was ultimately voice-driven, visual feedback remained essential. The prototype emphasized how visual cues could complement audio interactions, making the system feel more responsive and trustworthy.
Reflections & Next Steps
This project challenged me to design for emerging technologies while prioritizing clarity and user trust.
Key Lessons Learned
✔ Designing for voice requires visual thinking – Good VUI (voice user interface) design still relies on visual context to support user comfort and clarity.
✔ Familiar patterns build trust – Users respond well to cues modeled after successful consumer voice products.
✔ Visual prototyping helps de-risk innovation – Creating screen-based visualizations allowed stakeholders to understand and test the invisible.
Looking forward, this system could expand into multilingual support, staff training tools, or even hybrid kiosks for high-traffic environments—all while maintaining the same accessible, voice-first design foundation.
Prototype Demo Video
Please enjoy the prototype below to see the solution in action.
Click here to request a copy of the Figma file.