AI DevOps Automation Platform
Designing a training-free infrastructure automation system for engineers and cross-functional teams
The Outcome
The redesign significantly improved usability and operational efficiency.
Key outcomes:
• faster onboarding for DevOps engineers
• reduced deployment complexity
• improved troubleshooting workflows
• clearer configuration guidance
• scalable foundation for AI-driven DevOps automation
What began as a complex DevOps interface became a structured workflow engineers could use confidently without training.
50% Faster
Service creation time reduced through a guided workflow.
Simplified Forms
Complex configuration forms were redesigned into structured sections.
Training-Free UX
Guided flows reduced the need for onboarding.
Confidential
Confidential
Due to client confidentiality, real product screens and internal architecture cannot be displayed. This case study focuses on UX thinking, product decisions, and problem-solving.
The Context
This project involved redesigning a low-code DevOps automation platform used by engineering teams to deploy and manage cloud infrastructure.
The platform allows DevOps engineers to configure services, manage environments, monitor deployments, and troubleshoot system issues.
While the system was powerful, several workflows were difficult to use and required significant manual effort.
The goal of the redesign was to simplify complex infrastructure workflows while maintaining the flexibility required by DevOps engineers.
My Role
Worked as a UX & UI designer on the platform.
Responsibilities included:
• user research with DevOps engineers
• mapping infrastructure workflows
• redesigning the Add Services flow
• simplifying configuration forms
• designing an AI-assisted helpdesk interface
• improving observability dashboards
• collaborating with engineers and product managers
The Challenge
Despite strong automation capabilities, several workflows created friction.
Key issues included:
• steep learning curve for new engineers
• complex Add Services configuration flow
• fragmented troubleshooting experience
• observability dashboards difficult to interpret
• inconsistent form structures across modules
Engineers spent significant time configuring services, validating deployments, and troubleshooting errors.
Research & Discovery
To understand real workflows, interviews were conducted with DevOps engineers and solution architects who regularly manage customer environments.
Key insights
• engineers manage 40-50 services per environment
• onboarding often involves manual infrastructure data
• forms contained too many mandatory fields
• configuration mistakes frequently required rework
• validation feedback was unclear
One insight appeared consistently:
The Add Services form was extremely long and time-consuming.
Focus Area: Add Services Flow
Adding a service was one of the most important workflows in the platform.
However, the existing experience required engineers to complete a large configuration form with many technical parameters.
Problems included:
• unclear required vs optional inputs
• complex field dependencies
• validation appearing only after submission
• frequent configuration errors
Add Services Flow Redesign
The workflow was redesigned into guided configuration stages.
Instead of filling a single long form, users move through structured steps.
Improvements
• reduced redundant inputs by ~50%
• grouped fields into logical sections
• introduced smarter default configurations
• added inline validation and contextual guidance
• improved progress visibility
This approach helped engineers configure services step-by-step instead of navigating a dense form.
Observability Dashboard Redesign
The original observability module relied heavily on raw monitoring data.
While powerful, it was difficult for engineers to quickly interpret system health.
Observability Dashboard Redesign
The original observability module relied heavily on raw monitoring data.
While powerful, it was difficult for engineers to quickly interpret system health.
Improvements
• reorganized information hierarchy
• introduced summarized health indicators
• highlighted anomalies visually
• simplified navigation to deeper diagnostics
The redesigned dashboard provided a clear overview of service health and performance.
AI Helpdesk
To assist engineers during troubleshooting, an AI-assisted helpdesk interface was introduced.
Instead of leaving the platform to search documentation, users can now:
• request troubleshooting guidance
• ask for deployment fixes
• trigger configuration changes
• initiate rollback actions
The assistant provides context-aware responses based on the user’s environment.
AI Deployment Agents
The platform later introduced a broader vision: AI agents capable of executing DevOps workflows.
The idea was to create AI-powered “engineers” that assist with deployment and operational tasks.
System Structure
Engineer → Persona → Projects → Plans → Stages → Tickets → Execution
UX Approach
To simplify this system:
• AI engineers represent specialized agents
• personas define capabilities such as deployment automation or troubleshooting
• projects organize infrastructure tasks
• plans and stages break complex work into manageable steps
Execution happens through a chat-based interface, allowing engineers to interact with AI agents naturally.
Key UX Improvements
The redesign improved several aspects of the platform.
• Reduced cognitive load through guided workflows
• Improved transparency with clearer system feedback
• Safer configurations using smart defaults
• Actionable observability insights instead of raw monitoring data
• Contextual AI assistance embedded in workflows
Key Learnings
Designing for complex infrastructure systems reinforced several lessons:
• automation must remain transparent and explainable
• even expert users benefit from simplified workflows
• progressive complexity works better than exposing everything at once
• close collaboration with engineers is essential
Good UX in DevOps tools is not about removing power it is about making complexity manageable.