Agentic VM Migration

ROLE
UX Architect
Cross-Functional Lead

TEAMS
HPE - CTO Office, HPE GreenLake Cloud Platform
Private Cloud & Morpheus, Professional Services

Situation

I started with zero technical specs - just a high level vision from a cross-bu brainstorm session with Pm leaders

Task

Define the "Why"

Validate the use case with customers

Create a proof of concept

Action

Move from "Technology in search of a solution" to a value-driven product. Align cross functional teams across 2 orgs and 4 geos. Iterate in a fast paced, time boxed manner

Results

4 Time Zones Coordinated

98% VM Dependencies Mapped

Innovation Demo for HPE Discover

1 Patent Submitted

1 New VM Migration Service proposed on GreenLake Cloud roadmap

Discovery Phase

Customers

Home Depot, HPE Managed Services
While automation has helped to reduce the Migration planning and time taken, there is still a lot of manual effort involved in Discovery and Risk Assessment phases.

Key takeaways:

High-risk enterprise migration decisions
AI capability > human trust
Opaque automation slows adoption

Fragmented migration tooling, lengthy manual steps and assessments

The Pivot

We realized that users didn't want the AI to do the migration immediately; they wanted the AI to earn their Trust and prove it understood the environment first.

User Research

I led the technical and product teams in prioritizing agentic use cases based on business opportunity and biggest customer pain point.

Scope Prioritization

I led the technical and product teams in prioritizing agentic use cases based on business opportunity and biggest customer pain point.

Market Analysis

We compared some of the leading market solutions like AWS Transform, Google Cloud Migration Center, and Microsoft Azure Migrate for pros and cons about AI design patterns like explainability, shared agency along with assessment outputs.
This led to a deeper understanding of usability failures and need for AI principles and design patterns. Instead of static readiness reports, agents explain why a VM is risky, what dependencies matter, and where human judgment is required, increasing trust and adoption in complex hybrid environments.

AWS Transform
Link

Google Cloud Migration
Link

Microsoft Azure

Link

Gave us a general understanding of what industry type our potential clients belong too, responsibilities, motivations and pain points and best practices across existing market solutions.

EXISTING HPE Migration Tools

I organized discussions across Morpheus, HPE Private Cloud, HPE GreenLake Platform and HPE Professional Services teams in timezones to come up with priorities, customer and business value and key focus for this POC.

HPE Morpheus

Self-service, cloud based, limited functionality, SMB focus

HPE Rapid Migration Tool

On-prem, Managed Services: Home Depot, Danfoss etc

IDEATION PHASE

AI-Design Principles

Human-in-the-Loop by Default
Agents propose and act, but humans retain authority at key decision and escalation points.

Explain Before You Automate
Every recommendation includes the why, evidence, and assumptions—never outcomes without context.

Progressive Disclosure of Intelligence
Surface simple answers first, with deeper reasoning, data, and logs available on demand.

Confidence, Not Certainty
Agents communicate uncertainty through confidence scores and risk indicators, not false precision.

Learn from Human Overrides
Human corrections are signals, feeding continuous improvement and better future recommendations.

INITIAL CONCEPTS

Vibe Coding: Along with Engineering team, we explored several fast ideas for agents in VS Code, Figma Make etc . My role was to decide what would work for our users and pick the path forward.

AI Design Patterns

While these were early ideas for HPE Discover conference for the innovation booth, it led to development of AI patterns for discovery audit, confidence score and progressive disclosure for human-agent collaboration. See AI Patterns

Discovery Audit Card

A compact agent observability component that makes autonomous AI discovery visible, inspectable, and trustworthyin real time.

Confidence Score Badge

A lightweight agent observability component that makes AI uncertainty explicit by attaching a clear confidence signal to every AI-generated insight, recommendation, or decision.

Progressive Disclosure

A structured information layering pattern that reveals AI system detail in intentional levels allowing users to access transparency without overwhelming them.

Please contact me to view more details of the project and interactive prototype due to NDA

FEEDBACK RECEIVED

I though this was great - got a lot of people exposed to what is possible with agentic capabilities and how we can transform our workflows while building trust, safety and confidence. it gives a since of purpose and excitement to engineers and labs to focus their energy.

Chris Tinker, CTO, HPE Professional Services

Reflection

"When I took on the HPE GreenLake initiative, I didn't have a roadmap, I had to build one. I brought together fragmented teams from 4 different product silos and four time zones and to answer a fundamental question: Does this need an Agent, or just a better script? By focusing on Discovery and Risk Assessment, we addressed the 'Human-Trust Gap'. We used Vibe coding and Figma to move fast, but we used the Trust Evolution Model to ensure we were building something safe."

Page updated

Report abuse