Microsoft Azure Copilot Command Line Integration

Overview

The AKS Run Command project reimagined how developers interact with Kubernetes clusters by enabling command execution directly within the Azure Portal. I led the design of a new experience that integrates Copilot into the Run Command blade, helping users write and execute commands faster and with fewer errors.

Traditional workflows required developers to switch between the CLI, documentation, and the Azure Portal, increasing cognitive load and the chance of syntax mistakes. This project aimed to reduce that friction, simplify the command-execution process, and make the portal a smarter, more supportive environment for both new and experienced engineers.

Design Goals

• Integrate Copilot directly into the Run Command interface to provide real-time syntax assistance, examples, and intelligent suggestions

• Reduce command errors and task completion time

• Support both novice and expert users through progressive guidance

• Explore scalable design patterns for Copilot across other AKS workflows, such as YAML editing and cluster configuration

My Role: Lead Product Designer
I defined the interaction model and visual structure for Copilot within the Run Command blade. My work included wireframing, prototyping, and partnering with engineering to ensure feasibility and performance. I also facilitated design reviews with product and research teams, aligning design direction with technical constraints and user goals.

Research & Iteration

We conducted a focused usability study to understand how developers used the Copilot-enhanced Run Command interface. Participants performed common AKS tasks with Copilot’s inline guidance.

Early results showed that users loved the syntax help and examples but struggled with the separation between command drafting and execution. Based on these insights, I redesigned the interface to unify input, suggestions, and execution into a single flow. The update also introduced progressive guidance and clearer visual hierarchy, making it easier for users to discover and apply Copilot suggestions without leaving context.

Through multiple design iterations and heuristic reviews, the experience evolved into a cohesive, AI-assisted workflow that strikes a balance between flexibility and clarity.

Impact

• Launched the first Copilot-assisted Run Command experience in the Azure Portal

• Reduced time to execute standard AKS commands by improving accuracy and reducing context-switching

• Helped scale design patterns for future Copilot integrations across Azure workflows

• Empowered developers to move from task execution to insight and learning, making command-line operations more intuitive and human-centered

Reflection

This project reinforced my belief that the best AI integrations are not about automation alone, but about empowering people to work with more confidence. Designing for Copilot required balancing intelligence with transparency and giving users support when they need it, without ever getting in their way.