Overview
The enterprise application design process is rife with siloed knowledge, suboptimal decisions, and expensive rework. BluePrint makes designing applications easier and faster by recommending standardized patterns to meet developers' requirements.
I led design for the platform's minimum viable product, launched in Q4 2025.
Problem
Developers lack an easy way to discover and reuse standardized architectural patterns.
Through empathy interviews with developers and application architects, my team identified significant challenges in the enterprise application design process.
Developers often lack expertise in evolving technology trends and established architectural guidelines, and architects with this expertise are often consulted late in the process — after developers have already made redundant or suboptimal decisions. While a catalog of approved, standardized patterns for architecture exists, it is not well-known nor is adoption enforced.
The result: tech debt, expensive rework, and delayed project timelines.
Design process
From user flow to high-fidelity prototype.
Building upon requirements set forth by my product team, I ideated on the design of a platform that takes application intent and requirements and provides suggested architectural patterns. I evolved the concept from a user flow diagram to a high fidelity prototype.
By mapping the happy path, I identified missing functional requirements — including the critical need for save functionality so developers could leave and return without losing progress.
Sketching the UI fleshed out uncertainties in the platform's core capabilities. Drawing inspiration from Gemini and ChatGPT to ease a future transition to a conversational interface.
After stakeholder feedback, I created a high fidelity prototype using Material UI components to visualize the flow from login through receiving a recommendation.
User research
Moderated usability testing with 6 target users.
To evaluate the proposed UI design and information architecture, I conducted moderated usability testing with associates familiar with app architecture and/or with expressed interest in BluePrint:
- 1 cloud policy engineer
- 3 software engineers
- 2 architects (1 solutions, 1 application)
Using the insights I uncovered, my tech team was able to implement necessary improvements prior to MVP launch.
Research goals
Should the recommendation or chat log be presented first in previous questionnaires? Do users understand where the in-progress questionnaire alert will send them if they have multiple in-progress?
Insights — wins
The interface behaved as participants expected.
For the most part, participants appreciated BluePrint's intuitive design and recognized its potential for usefulness in their day-to-day work with application architecture.
Participants knew how to start a questionnaire and where to find past questionnaires without prompting.
Users expected dynamic, guided questions about their application and what they were trying to achieve — and that's exactly what they got.
The contents of the recommendations — pros and cons, a component diagram, multiple patterns to compare — aligned with participant expectations.
Insights — MVP improvements
All improvements were content-related.
Of the few improvements needed for the MVP design, none were structural — all were about what the product said, not how it was shaped.
Clearer explanation of BluePrint's capabilities and value proposition on the login screen.
Present output recommendations before input response when viewing a past questionnaire.
Make explicit that the in-progress banner will take users to their most recent unfinished questionnaire.
"Generic — I don't understand how it will help me."
Insights — enhancements
Opportunities that shaped the product roadmap.
Multiple opportunities to enhance BluePrint's functionality were identified during testing. Most participants were confused by the JSON output format — what it represented and what they could do with it. They wanted something they could use "right away" and play around with.
Other desired capabilities initially contradicted requirements set forth for the MVP by my product partner — editing, deleting, and sharing questionnaires. The team believed these functionalities could allow users to "game" the recommendation engine. However, since BluePrint is a tool and not law, I was able to advocate for these features, and the insights I surfaced helped influence the product roadmap and shift the engineering team's focus.
- Output file format that can be more easily shared and deployed (e.g., vs. JSON)
- Edit and delete functionality for questionnaire responses
- Shareable links to BluePrint questionnaires and recommendations
Looking forward
A well-received MVP with clear direction for what comes next.
Overall, BluePrint's MVP was well-received by users — they recognized the platform's potential to streamline the architectural design process, especially once AI was integrated. As the focus for the next iteration turned to AI integration, the user research I conducted gave the team a clear and prioritized enhancement roadmap.
My personal learnings centered on collaborative habits: including my product partner during synthesis would have provided richer context earlier and reduced the friction of advocating for user-driven changes later.