Agent-Assisted Change Management Platform
UX/UI — AI Workflow Design
Designed an AI-assisted workflow platform integrated with ServiceNow to guide engineers through complex change management processes using coordinated agents.
Problem
Enterprise change management workflows are:
- complex and multi-step
- time-consuming and error-prone
- difficult to navigate without deep expertise
There was an opportunity to:
- augment this process using AI agents
- reduce cognitive load
- guide users through decisions and validation steps
My Role
- Led UX/UI design for early-stage product concept (0 → proof of concept)
- Designed end-to-end user flows for agent-assisted workflows
- Acted as a bridge between product ambition and engineering feasibility
- Delivered high-fidelity prototypes to communicate product vision
Key Contributions
1. End-to-End Agent Workflow Design
- Designed a multi-step change management flow guided by specialized agents
- Structured interactions where agents analyze intent, assess impact, generate test plans, and execute and validate changes
- Shifted user role from executor → reviewer/decision-maker
2. Timeline-Based Interaction Model
- Created a timeline interface allowing users to navigate across stages of the change process and revisit and validate previous steps
- Improved clarity and control in a complex, multi-phase workflow
3. Multi-Agent Interaction System
- Defined interaction patterns for multiple agents (planner, tester, analyst, etc.)
- Designed conversational + structured UI hybrid (chat + system actions)
- Enabled coordination between agents while maintaining user oversight
4. Data Visualization & System Context
- Designed network topology visualization for system awareness
- Enabled users to explore, search, and understand infrastructure impact
5. Document & Output Interaction
- Designed expandable, interactive outputs (test plans, reports, summaries)
- Allowed users to review, validate, and iterate on generated artifacts
6. Dashboard & Workflow Management
- Designed overview dashboard for active change requests, status tracking, and key updates and outputs
7. Cross-Functional Alignment
- Acted as a mediator between product and engineering
- Translated ambitious product vision into feasible design solutions
- Balanced innovation with technical constraints
Impact
- Delivered a full proof-of-concept for an agent-driven enterprise workflow
- Established foundational interaction patterns for AI-assisted operations
- Design direction was carried forward into subsequent iterations of the product
More projects
Design System for Enterprise Products
Open-source design system to support multiple emerging products within Cisco’s innovation incubator (Outshift), enabling consistency, speed, and adaptability across ventures.
AI Assistant for Enterprise Product
Improving an AI assistant through user research, UX/UI, and interaction design. Exploring how contextual interactions can make AI more intuitive and integrated.