Nova AI
Personal Project
Role
Product Designer
Duration
2 Weeks
Project
Project
UI/UX Design
Human AI Interaction
Human AI Interaction
In today’s digital landscape, many interfaces rely heavily on chat-based interactions. While this offers users flexibility, it often creates challenges around effective communication and setting the right context. Users may struggle to express their needs clearly or navigate overly complex conversations. In this project, I focused on designing an experience that strikes a balance between flexibility and simplicity — helping users communicate more effectively while keeping the interaction intuitive and purposeful.
In today’s digital landscape, many interfaces rely heavily on chat-based interactions. While this offers users flexibility, it often creates challenges around effective communication and setting the right context. Users may struggle to express their needs clearly or navigate overly complex conversations. In this project, I focused on designing an experience that strikes a balance between flexibility and simplicity — helping users communicate more effectively while keeping the interaction intuitive and purposeful.
Communication
Large language models (LLMs) typically start without any embedded context, placing the burden on users to provide detailed prompts.
Large language models (LLMs) typically start without any embedded context, placing the burden on users to provide detailed prompts.
This makes it difficult to fully unlock their potential for nuanced, context-driven tasks. As a result, AI is often seen merely as a tool for generating outputs, rather than as an active collaborator. In this project, I explored how better interface design — balancing flexibility with simplicity — can help users communicate more effectively, set richer context, and work more seamlessly with AI.
This makes it difficult to fully unlock their potential for nuanced, context-driven tasks. As a result, AI is often seen merely as a tool for generating outputs, rather than as an active collaborator. In this project, I explored how better interface design — balancing flexibility with simplicity — can help users communicate more effectively, set richer context, and work more seamlessly with AI.
Without built-in constraints or contextual guidance, users may struggle to set up the right context, leading to longer iteration cycles and less effective outputs.
Without built-in constraints or contextual guidance, users may struggle to set up the right context, leading to longer iteration cycles and less effective outputs.
This project explores how rethinking interaction patterns can help users communicate more effectively with AI, striking a balance between flexibility, simplicity, and intuitive guidance.
This project explores how rethinking interaction patterns can help users communicate more effectively with AI, striking a balance between flexibility, simplicity, and intuitive guidance.
How Might We
HMW connect users with AI to maximize its abilities and minimize the effort needed to set context?
HMW connect users with AI to maximize its abilities and minimize the effort needed to set context?
Ideate
Chat-based interfaces offer users flexibility and a familiar way to interact, requiring minimal learning to get started.
Chat-based interfaces offer users flexibility and a familiar way to interact, requiring minimal learning to get started.
However, unlike traditional “noun-then-verb” mechanics — where users first select an object and then apply an action — chat interfaces often lack the clear structure that makes interactions intuitive and efficient. Without clear constraints or guidance, users may struggle to set the right context, leading to longer iteration cycles and less effective outputs.
However, unlike traditional “noun-then-verb” mechanics — where users first select an object and then apply an action — chat interfaces often lack the clear structure that makes interactions intuitive and efficient. Without clear constraints or guidance, users may struggle to set the right context, leading to longer iteration cycles and less effective outputs.
In this project, I explored how adding constraints to the action space can help users narrow their focus, set clearer context, and enable AI systems to generate more accurate and relevant outputs — ultimately balancing flexibility, simplicity, and guidance to create a more effective communication experience.
In this project, I explored how adding constraints to the action space can help users narrow their focus, set clearer context, and enable AI systems to generate more accurate and relevant outputs — ultimately balancing flexibility, simplicity, and guidance to create a more effective communication experience.
By introducing constraints into the action space, we can help users narrow their focus and define clearer context for the AI.
By introducing constraints into the action space, we can help users narrow their focus and define clearer context for the AI.
This not only reduces cognitive load, but also enables the AI to generate more accurate and relevant outputs. Instead of leaving the interaction entirely open-ended, lightweight guidance—such as suggested actions, filters, or contextual cues—can empower users to communicate more effectively while still preserving the flexibility of a chat-based interface.
This not only reduces cognitive load, but also enables the AI to generate more accurate and relevant outputs. Instead of leaving the interaction entirely open-ended, lightweight guidance—such as suggested actions, filters, or contextual cues—can empower users to communicate more effectively while still preserving the flexibility of a chat-based interface.
Interaction
Users should be able to easily and intuitively specify what they’re referring to in their prompts.
Users should be able to easily and intuitively specify what they’re referring to in their prompts.
y allowing users to anchor their input to specific objects, actions, or contexts, we introduce subtle constraints that guide the AI’s understanding. This added structure helps narrow the action space, enabling the AI to generate more accurate, relevant, and context-aware outcomes—without sacrificing usability or flexibility.
y allowing users to anchor their input to specific objects, actions, or contexts, we introduce subtle constraints that guide the AI’s understanding. This added structure helps narrow the action space, enabling the AI to generate more accurate, relevant, and context-aware outcomes—without sacrificing usability or flexibility.
An effective AI assistant should integrate seamlessly into the user’s existing workflow, allowing interaction without requiring users to switch contexts or leave their current environment.
An effective AI assistant should integrate seamlessly into the user’s existing workflow, allowing interaction without requiring users to switch contexts or leave their current environment.
By co-existing in the same space, the AI becomes a natural extension of the workspace—reducing friction, enhancing efficiency, and providing support directly where tasks are happening. This approach not only streamlines the experience but also helps AI feel more like a collaborative partner than an external tool.
By co-existing in the same space, the AI becomes a natural extension of the workspace—reducing friction, enhancing efficiency, and providing support directly where tasks are happening. This approach not only streamlines the experience but also helps AI feel more like a collaborative partner than an external tool.
Better interaction patterns can help users provide more specific input, improving how effectively AI understands and responds.
Better interaction patterns can help users provide more specific input, improving how effectively AI understands and responds.
Point & Select: Let users clarify intent by selecting or highlighting specific elements in their workspace—like circling an area, selecting a text block, or pointing to a UI component.
Point & Select: Let users clarify intent by selecting or highlighting specific elements in their workspace—like circling an area, selecting a text block, or pointing to a UI component.
Contextual Menus: Introduce dynamic, context-aware menus (similar to a “right-click” experience) that surface relevant actions based on the user’s selection.
Contextual Menus: Introduce dynamic, context-aware menus (similar to a “right-click” experience) that surface relevant actions based on the user’s selection.
Better interaction models can help users express intent more precisely, enabling AI to deliver more relevant results.
Better interaction models can help users express intent more precisely, enabling AI to deliver more relevant results.
Flexible Tools: Offer tools like plugins or overlays that expand into full-screen mode (similar to Loom), allowing users to interact with AI without breaking focus or leaving their current workspace.
Flexible Tools: Offer tools like plugins or overlays that expand into full-screen mode (similar to Loom), allowing users to interact with AI without breaking focus or leaving their current workspace.
Platform Integration: Leverage familiar interaction patterns like object naming, selection, and grouping found in tools such as Figma, Photoshop, or Illustrator to help users refer to specific elements with greater accuracy.
Platform Integration: Leverage familiar interaction patterns like object naming, selection, and grouping found in tools such as Figma, Photoshop, or Illustrator to help users refer to specific elements with greater accuracy.
By aligning with existing workflows and tools, these enhancements reduce ambiguity and make AI collaboration feel more natural and efficient.
While flexibility unlocks powerful functionality, it often comes at the expense of simplicity.
While flexibility unlocks powerful functionality, it often comes at the expense of simplicity.
To bridge this gap, the interface should support progressive onboarding and ease of discovery—enabling users to gradually learn and master features without feeling overwhelmed.
To bridge this gap, the interface should support progressive onboarding and ease of discovery—enabling users to gradually learn and master features without feeling overwhelmed.
By layering complexity and offering just-in-time guidance, we can design experiences that are both intuitive for new users and flexible enough for advanced use cases.
By layering complexity and offering just-in-time guidance, we can design experiences that are both intuitive for new users and flexible enough for advanced use cases.
Solutions
Reducing Friction for First-Time Users
Reducing Friction for First-Time Users
To make the AI assistant more approachable and intuitive, the interface provides lightweight guidance right from the start:
To make the AI assistant more approachable and intuitive, the interface provides lightweight guidance right from the start:
Clear Starting Point: The placeholder text “Get started with anything you need assistance with” sets expectations and removes ambiguity, making it easier to initiate interaction.
Clear Starting Point: The placeholder text “Get started with anything you need assistance with” sets expectations and removes ambiguity, making it easier to initiate interaction.
Suggested Actions: Quick prompts highlight common use cases, helping users understand the assistant’s capabilities while lowering the barrier to engagement.
Suggested Actions: Quick prompts highlight common use cases, helping users understand the assistant’s capabilities while lowering the barrier to engagement.
Together, these elements reduce the learning curve, build confidence, and make the AI assistant feel more accessible—especially for new users.
Together, these elements reduce the learning curve, build confidence, and make the AI assistant feel more accessible—especially for new users.
Solutions
Easy to Access
Easy to Access
The AI assistant is designed to be readily available across the user’s workflow, minimizing friction and maximizing convenience:
The AI assistant is designed to be readily available across the user’s workflow, minimizing friction and maximizing convenience:
Universal Accessibility: Users can summon the assistant from any window or screen—whether they’re working in a document, design tool, or browser—without needing to switch apps or break focus.
Universal Accessibility: Users can summon the assistant from any window or screen—whether they’re working in a document, design tool, or browser—without needing to switch apps or break focus.
Persistent Entry Point: A floating icon or keyboard shortcut ensures that help is always just a click or tap away, encouraging ongoing use and making the assistant feel like a natural part of the workspace.
Persistent Entry Point: A floating icon or keyboard shortcut ensures that help is always just a click or tap away, encouraging ongoing use and making the assistant feel like a natural part of the workspace.
By embedding the AI directly into the user’s environment, the experience feels more seamless, contextual, and supportive—wherever and whenever it’s needed.
By embedding the AI directly into the user’s environment, the experience feels more seamless, contextual, and supportive—wherever and whenever it’s needed.
Solutions
The Right-Click Approach
The Right-Click Approach
Bringing familiar interaction patterns into AI-driven workflows helps users express intent more precisely and intuitively:
Bringing familiar interaction patterns into AI-driven workflows helps users express intent more precisely and intuitively:
Context-Specific Actions: When users right-click (or tap-and-hold) on an object, the system surfaces context-aware options directly related to the selection. This helps narrow the scope of intent without requiring users to describe it manually in a prompt.
Context-Specific Actions: When users right-click (or tap-and-hold) on an object, the system surfaces context-aware options directly related to the selection. This helps narrow the scope of intent without requiring users to describe it manually in a prompt.
Interactive and Intuitive: This approach mirrors traditional “noun-then-verb” workflows—selecting an object before applying an action—which users already understand. It makes AI interaction feel more natural and aligned with existing behavior patterns.
Interactive and Intuitive: This approach mirrors traditional “noun-then-verb” workflows—selecting an object before applying an action—which users already understand. It makes AI interaction feel more natural and aligned with existing behavior patterns.
By combining AI with familiar UI mechanics, we reduce friction, improve accuracy, and make the system feel more predictable and user-friendly.
By combining AI with familiar UI mechanics, we reduce friction, improve accuracy, and make the system feel more predictable and user-friendly.
Solutions
Specify Focus Area
Specify Focus Area
Enabling users to visually define the focus area enhances clarity and improves the AI’s contextual understanding:
Reduce Misinterpretation: By selecting or highlighting a specific section—whether it’s a block of text, a UI element, or part of a design—users can directly indicate what they want the AI to work with. This minimizes ambiguity and reduces the risk of irrelevant or incorrect responses.
Faster Task Completion: Visual selection removes the need for long, detailed prompts. It shortens the back-and-forth and helps users reach accurate results more quickly and efficiently.
This approach not only improves output quality but also creates a more seamless and user-controlled interaction experience.
Embed AI Within the Workspace
Embed AI Within the Workspace
Integrating the AI assistant directly into the user’s workspace allows for more seamless, context-aware collaboration. Instead of requiring users to switch tools or environments, the assistant remains accessible wherever work is happening—reducing friction and maintaining focus.
Integrating the AI assistant directly into the user’s workspace allows for more seamless, context-aware collaboration. Instead of requiring users to switch tools or environments, the assistant remains accessible wherever work is happening—reducing friction and maintaining focus.
By adding subtle constraints to user input (e.g., through selection, context menus, or embedded tools), the system can better interpret intent and scope its responses accordingly. This combination of in-context access and focused input leads to more accurate results, faster task completion, and a smoother overall experience.
By adding subtle constraints to user input (e.g., through selection, context menus, or embedded tools), the system can better interpret intent and scope its responses accordingly. This combination of in-context access and focused input leads to more accurate results, faster task completion, and a smoother overall experience.
Chat Interface
Chat-based interfaces offer users flexibility and a familiar way to interact, requiring minimal learning to get started.
Chat-based interfaces offer users flexibility and a familiar way to interact, requiring minimal learning to get started.
However, unlike traditional “noun-then-verb” mechanics — where users first select an object and then apply an action — chat interfaces often lack the clear structure and affordances that guide user behavior.
However, unlike traditional “noun-then-verb” mechanics — where users first select an object and then apply an action — chat interfaces often lack the clear structure and affordances that guide user behavior.
Let’s Talk
© 2025 Nutcha. All Rights Reserved.