Adobe
During my summer internship at Adobe, I developed an AI-assisted onboarding agent for Adobe Experience Manager, utilizing Edge Delivery Services. This product is designed for individuals who lack technical experience and have no time to spare.
*Proprietary work completed during Adobe internship.
All rights reserved by Adobe Inc.
Duration: 12 Weeks
Role: Designer developing a new AI-assisted onboarding flow
Tools: Figma, Miro, V0, Cursor, and Adobe Creative Cloud
Deliverables: High-Fidelity Wireframes
Currently, migrating or launching sites on Adobe’s Edge Delivery Services (EDS) requires relying on Adobe’s internal teams for support. This process slows down timelines, creates bottlenecks, and prevents business users from working independently, ultimately increasing TCO.
In simpler words...
Current Model
Need of a Technical Team (5+ people)
Inefficient
(2-3 months)
Costly
Challenging
New Model
A one Man Show!
Rocket Speed
(A couple of hours)
Less time= Less Resources
A Happy, NOT Stressed, User
Which after 12 weeks looked like this
1
Scan & Analyze Your Site
Step 1:
Data analyst agent
beings site scan
Step 2:
Get a generated
summary of your site
Step 3:
Expand document to
learn more about your
sites metrics
2
Managing Oppurtunities
Step 1:
Get a generated list of possible opportunities
Step 2:
Filter opportunities by metric or page
Step 3:
Complete selection and wait for updates
3
Migration Recommendations
Step 1:
Review the three recommended plans
Step 2:
Compare plans you're curious about
Step 3:
Analyze the two plans to make a decision
4
Migration Review
Step 1:
Pick your plan and continue with migration
Step 2:
Review migration outline and updated assets
Step 3:
Wait for approval and try a new task
Let's start from the beginning, who am I designing for?
Meet
Ananya
An E-commerce Manager @ WKND
Sarah has recently been assigned to lead a site migration project into AEM. While she’s new to the technical aspects of site migrations, she views this migration as an opportunity not only to move content but also to streamline the site’s structure and enhance overall performance.
Her Technical Knowledge
Her Technical Experience
What are her goals?
To always get green lightehouse scores
To customize every step of the migration process
To complete migration with ease within hours not days
How is the current journey affecting Ananya?
Lack of Control
Involvement of Adobe’s internal teams limits customer autonomy, causing it to be time-consuming and costly, often surpassing the cost of the license itself.
Technical Dependency
The migration process is complex, resulting in a steep learning curve for technical and non-technical users.
Absence of Personalization
Adobe’s internal teams work on various projects, leading to standardized design, not leaving much room for customization.
Analyzing the Market
After understanding the user, it was time to understand content management systems (CMS), site migration, AI industry trends, and features and experiences already in use.
The best way to create something that lasts is to know what works and what doesn't; that's why, for this competitor analysis, my main goal was to explore experiences and user flows that work and identify the pain points behind those that don't.
Miro board competitor analysis highlighting insights from 8 different competitors
Here is a breakdown of my approach:
-
I needed to address both Content Management systems and onboarding platforms, so I decided to take a half-and-half approach and examine both groups (onboarding and site migration tools) together.
-
Some of the metrics I used were how AI is involved, similarities/differences to AEM, features for future implementation, how automated/lengthy the process is, and more
.png)
Here is a closeup of the 8 main competitors I looked at
Follow-up AI Research
Using the findings from the competitor analysis, I looked into how AI plays a role in human-centered experiences and when to draw the line in heavily automated user flows. Here is a list of topics I focused on:
Identifying Common AI Modalities and inspirations
Industry trends, AI use cases+ impact, unique features/tools
Discovering patterns within AEM
Whats been done before and why, any pain points?
Consider similarities within AI and UX Patterns
Is there a way to blend the two flows? Will it create a higher impact?
.png)
Creating human–centric and conscious designs
Why do we use AI? How can we build user trust? What are the positives and negatives of different patterns?
Here are the key takeaways
AI Personalization
Making the experience as personalized as possible, from the type of agent to focused optimizations.
AI Transparency
Making AI more accessible and
informative about its next steps helps the user trust the agent and its responses more
AI with a Human Touch
Making the AI agent intuitive and matching the human line of thought, as well as constantly learning from a user's actions
A design ideology that later guided me in my design process
During AI’s formative stages, UX designers must leverage dynamic visuals and multisensory expression to bridge the gap between AI’s invisible power and its practical, everyday utility.
- Ken Olewiler
Beginning the Design Process
Using the first week and a half of research, I began listing the goals I wanted this experience to accomplish. After grouping actions and processes, I identified 6 Main goals. I then ideated a bunch of features and opportunities I wanted to include, but there were way too many and too little time, so here's what I did:
-
Using the 6 main goals, I organized all features, user flows, and opportunities to be among the 6 goals.
-
Any feature that didn't find a home within the 6 goals was booted.
-
From there, I made a feature prioritization chart and created a timeline to work on those that are high priority first and make my way down the list.
The 6 Main Goals
Self- Sufficiency
Equip business users with a no-code onboarding flow that allows them to migrate websites without requiring hands-on support
Simplify Migration
Automate the detection and migration of key content, structure, and features from legacy platforms
Optimize
Ensure that the onboarding experience incorporates built-in recommendations for performance, SEO, accessibility, and responsiveness
Build Confidence
Use explainable AI, contextual help, and progress tracking to give users the confidence to complete the migration.
Minimize Time-to-Value
Streamline onboarding to reduce the time from first interaction to live EDS deployment
Scale with Intelligence
Allow the system to learn from past migrations & continuously improve recommendations and simplify future onboarding.
Miro board showcasing all ideas organized by goal
Miro board showcasing all ideas organised by priority and impact
Based on these main features I conducted a small brainstorming session with my team.
Mine!
This lead to the newly designed user journey with 4 main steps:
1
Welcome and Setup
Set up account, onboard your team, upload site, personalize goals and optiizations
2
Scan and Integrate
Scan site, review metric analysis, redesign site map, select pages for migration
3
Migration Blueprint
Review migration recommendation, A/B test different plans, build custom workflow, assign tasks
4
Finish Setup
Final tutorial, setup dashboard, review team and project hubs, locate calendar
Starting with Lo-Fi's
Spoiler alert...the feedback wasn't great.
Here were some of the reviews:
...try to always have the easiest path for the user...right now the user is inputting information that the AI should already know and use- Manager
AI needs to be the first hand experience not second- Mentor
Refining the Journey
Using the new feedback I re-constructed the onboarding journey to this:
Scan & Analyze site
1
Manage Oppurtunities
2
Review Migration Plan
3
Choose Migration Plan
4
An expanded onboarding flow illustrating the main actions taken by the chat agent and user
...this leads us back to the final designs!
1
Scan & Analyze Your Site
Step 1:
Data analyst agent
beings site scan
Step 2:
Get a generated
summary of your site
Step 3:
Expand document to
learn more about your
sites metrics
2
Managing Oppurtunities
Step 1:
Get a generated list of possible opportunities
Step 2:
Filter opportunities by metric or page
Step 3:
Complete selection and wait for updates
3
Migration Recommendations
Step 1:
Review the three recommended plans
Step 2:
Compare plans you're curious about
Step 3:
Analyze the two plans to make a decision
4
Migration Review
Step 1:
Pick your plan and continue with migration
Step 2:
Review migration outline and updated assets
Step 3:
Wait for approval and try a new task
Next Steps & Lessons Learned
Im forever grateful for the opportunity to learn, grown and connect with so many talented individuals. Although my 12 week are up at Adobe, here are some of the next steps for this project:
Finalize Design Interacions
Sync all interactions together along with the dashboard
Explore Microinteractions
Look into notifications, offline activity and different agent modes
Collect User Feedback
Conduct user testing and redefine the flow based on feedback
Project Handover
Handover project to stakeholders and document all research and design files for easy use
this project taught me that...
You Can Only do so Much
With limited time, I did not have enough time to explore and build everything I wanted to
It's Hard to Gather Users
With an enterprise project its harder to find people to converse and understand pain points from
Change is In
Conduct user testing and redefine the flow based on feedback
Designing a Feeling is Hard
Handover project to stakeholders and document all research and design files for easy use