SP
AMS Reliability Advisor
The first AI-powered assistant for Emerson's AMS Device Manager ecosystem
Overview
Role
UX Designer
Timeline
August - November 2026
Tools
Figma
Claude Design
Microsoft Copilot
Team
2 UX Designers
Project Manager
Project Owner
Engineers
Background
AMS Device Manager is Emerson's software for monitoring, maintaining, and troubleshooting industrial field devices such as valves, transmitters, and sensors. For decades, it has helped maintenance and reliability teams manage critical equipment across industrial facilities.
As part of a broader modernization effort, Emerson redesigned AMS Optics, an existing web-based platform for monitoring device health, alerts, and reliability data. (see my AMS Optics redesign case study [here]) Alongside this initiative, a separate effort explored how AI could further enhance the user experience.
This led to AMS Reliability Advisor, the first AI-powered assistant within the AMS Device Manager ecosystem. Embedded directly in AMS Optics, it provides conversational support and helps users interact with the platform more efficiently.
Depending on a customer's subscription tier, the assistant could:
-
Answer questions and help users navigate AMS Optics (AI Agent)
-
Perform actions through guided workflows (AI Agent + Advisor)
Problem
AMS Optics introduced a completely new way for users to interact with AMS Device Manager data. Many existing customers had spent years working within familiar workflows and were not necessarily comfortable adapting to new digital experiences.
As the platform evolved, users needed a way to:
-
Learn unfamiliar workflows without extensive training
-
Find information quickly across large amounts of asset data
-
Complete tasks more efficiently
-
Build confidence while transitioning to a modernized experience
At the same time, AI assistants were becoming an increasingly common interaction model across software products. We saw an opportunity to use AI not simply as a help tool, but as an integrated experience that could guide users, answer questions, and eventually help them complete tasks directly within the product.
Project Goals
Create an approachable assistant that could simplify complex workflows while increasing user confidence in a newly modernized platform.
Discover
Ideate
Design
Validate
What problems should the AI Advisor solve?
Discover
Competitive Analysis
We analyzed conversational AI products such as ChatGPT, Microsoft Copilot, and Claude to identify common interaction patterns and user expectations.
Copilot
-
Prompt suggestions and templates
-
AI-generated actions and next steps
-
Integrate product workflows
-
Conversational assistance paired with task completion
ChatGPT
-
Suggested conversation starters
-
Chat history and conversation management
-
Context-aware responses
-
Support for both informational and task-based requests
Claude
-
Minimal, conversation-first interface
-
Suggested discovery prompts
-
Long-form contextual responses
-
Clear guidance and recommendations
Patterns Across Leading AI Assistants:
-
Conversational interfaces centered around a chat experience
-
Suggested prompts to help users get started
-
Context-aware responses with actionable next steps
-
The ability to ask both informational and task-oriented questions
-
Clear distinction between asking questions and performing actions
These patterns helped establish a foundation for designing an experience that would feel familiar and intuitive to users.
Stakeholder Discussions
We collaborated closely with project managers, product owners, and senior principal engineers to understand business goals, technical constraints, and user needs.
-
Assist users with product knowledge and navigation
-
Answer basic questions
-
Help users find information within AMS Optics
-
Guide users through workflows and tasks
2. Enable action-oriented assistance for Agent subscribers
-
Perform specific tasks on behalf of users
-
Automate selected workflows
-
Reduce manual effort for common actions
Through these discussions, we identified two primary objectives for the AI Advisor:
How might we design a familiar and intuitive AI experience?
Ideate
Key Insights
From our research, we identified three key opportunities:
1
2
3
Users expect AI interactions to follow familiar conversational patterns.
The AI Advisor needs to support both informational and action-based requests.
The experience has to accommodate users with different permission levels and capabilities based on their subscription.
Defining the Experience
We explored how a conversational experience could support both guidance and action by focusing on:
-
Helping users understand what the AI can do
-
Making available capabilities easy to discover
-
Supporting a variety of request types
-
Creating a clear path between asking questions and completing actions
How did we bring the AI Advisor experience to life?
Design
Wireframes
Building on insights from the discovery phase, we explored different ideas and refined them through several design iterations.
Standard View
.png)
Low - Fidelity
Mid - Fidelity
High - Fidelity
Expanded View
Full Screen View
.png)
Research-Informed Features
Discovery Finding → Design Decision
Highlighted Workflow:
Modify an Alert Configuration
Among the many scenarios supported by the AI Advisor, modifying an alert configuration was chosen as a representative happy path because it demonstrates both conversational guidance and task execution.
User selects a prompt.
User selects the alert they'd like to modify.
User adjusts the alert settings and reviews.
The AI summarizes the updates and asks for confirmation before applying them.
The AI confirms the configuration has been updated.
An interactive prototype is currently being finalized as the feature moves toward implementation.
How are we validating the solution?
Validate
Stakeholder Review & Handoff
The proposed designs were reviewed with subject matter experts (SMEs), project stakeholders, and engineering partners to ensure the experience aligned with user needs, business goals, and technical requirements.
Following several rounds of feedback and refinement, the designs were approved and handed off to the engineering team for implementation.
Expected Impact
AMS Reliability Advisor introduces a tiered AI experience through Advisor and Agent subscriptions. By offering advanced capabilities such as task execution and automation through higher-tier plans, the product creates new opportunities for monetization while delivering increased value to users.