2022
Kyle Laporte
Blaise
Military
AI
SaaS
Air Force personnel needed a fast, reliable, and intuitive way to test jet fuel quality in the field. Existing diagnostic tools were complex, lab-dependent, and time-consuming creating risks for mission readiness and equipment safety.
Overview
Blaise™ is a low-cost, portable, and AI-powered handheld Raman spectrometer that revolutionizes real-time detection of pathogens, early-stage cancer biomarkers, and chemical contaminants. Designed for both field use and clinical environments, Blaise™ integrates hyperspectral Raman spectroscopy, microfluidic chips, and machine learning algorithms to deliver lab-grade precision on the go.
My primary mission as Lead UX/UI Designer was to craft an intuitive, high-performance interface that empowers users—from Air Force personnel to biomedical technicians—to make rapid, informed decisions under pressure.
Project goals
1.
Clear, legible data visualization in harsh lighting conditions and compatibility with gloves, protective gear, and low-light scenarios.
2.
Minimal cognitive load during high-stress operations.
3.
Rapid onboarding and error-resistant workflows.
Challenge
The U.S. Air Force needed a handheld diagnostic tool capable of detecting jet fuel contamination and biological threats within seconds.
The challenge was twofold:
Designing for high-stakes environments where every second counts, and the UI must communicate complex data simply and clearly.
Translating lab-grade analytics into field-friendly insights ensuring users with varying technical expertise could operate the device confidently.
Research & Discovery
I began by conducting contextual inquiries and field interviews with Airmen and lab technicians to understand operational environments and cognitive workflows.
Simplicity over density: Users needed a “traffic light” style status indicator (e.g., Green = Safe, Red = Contaminated).
Task-based navigation: Instead of data-heavy screens, users preferred guided steps (“Scan → Analyze → Result”).
Feedback immediacy: Audio, haptic, and visual confirmations reduced uncertainty in critical situations.
How might we design a handheld interface that enables Airmen to perform rapid, accurate fuel and pathogen detection without requiring scientific expertise?
User Flow Mapping
I mapped the device workflow into a three-step process:
Scan – Initiate a scan with one button.
Analyze – Machine learning models process spectra in real-time.
Report – Results are displayed with clarity and recommended next steps.
Wireframes - Iterations
In the early iterations of the design many of the frameworks for the scanning feature had not been fully developed.
The machine learning system was still in development, The chemical analysis was still being finalized and little actual user research had been done at this point.
While, the science division continued testing I began initial concepts and tested with anticipated users frequently over 6 months.
The feedback and learnings from this research was invaluable to building the final product design.
Below is the journey of the main entry point to testing Flow took through rapid iteration and continued testing.
Testing Flow 1
My Initial concept
based on only the brief combined multiple flows while allowing for additional menus.
Advanced analysis and report indicator was initially all one screen
Advanced analysis and report indicator was initially all one screen
Testing Flow V2
From initial feedback I altered the designs to address two core issues:
A majority of the potential personas would likely be wearing gloves.
I opted for a more minimalistic approach that removed the spectrograph and enlarged the main feature points of access to accommodate for rugged use
Testing Flow v3
As development continued the team discussed including multiple scanning modes for different use cases
I incorporated progressive disclosure to nest the
laser exposure options and reduce cognitive load within the screen
User testing showed that we had a 9% engaged session rate on the V3.2 design - It wasn’t immediately clear whether this confusion stemmed from the testing screen options or the testing flow itself.
I decided to try an A-B test with similar starting flows to validate potential areas
Testing Flow v3.4 A
I simplified the layout to only one
Test type options and instead used
Tabs to transition between laser exposure
Testing Flow v3.4 B
The second option kept the test type relationship the same but allowed for laser exposure adjustments when the test began.
I also included recent tests as a horizontal scroll for quick access.
Continued testing showed more users gravitated towards version A, but the recent tests carousel was also of great interest
38%
of users interacted with the recent tests section despite lower overall engagement,
V 3.4 A
Click-through rate (CTR)
6.8%
Average session duration
2m 45s
Interaction with “Recent Tests” carousel
N/A
Return visit rate
24%
Preferred
V 3.4 B
Click-through rate (CTR)
5.2%
Average session duration
2m 10s
Interaction with “Recent Tests” carousel
38%
Return visit rate
22%
Additional qualitiative insights pointed towards a hybrid between versions a and b
“I like how clean and direct Version A feels I can switch through the tabs faster.”
“The recent tests carousel caught my eye. It’s nice to have that right there.”
“Version A feels more professional; Version B is more visually interesting.”
Testing Flow v5
The next iteration became an intentional blend of both versions incorporating the recent tests and keeping the main test panel
Testing Flow v6
After final polish:
Final Testing Flow
Outcomes
Designing Blaise™ challenged me to merge AI-driven complexity with mission-critical simplicity. It was an exercise in empathy crafting interfaces that empower Airmen and clinicians alike to trust technology when precision and time truly matter.
Reduced analysis time from minutes to seconds.
Enabled non-specialists to conduct complex fuel quality testing with confidence.
Improved operational readiness by minimizing maintenance downtime due to contaminated fuel.
30% more Users
Answers Hub
FAQ
Book a 15-min call
What kinds of projects do you take on?
I specialize in end-to-end mobile app development from idea
validation and UI/UX design to handoff. I will work with you and
your team to achieve a smooth deployment and include post-
launch support
Do you also code the apps you design?
I handle both design and development, employing lowcode and
What deliverables will I receive?
You’ll receive everything you need to launch and grow your app,
How do you price your work?
Our pricing is based on the scope and complexity of your
What’s the typical timeline?
Every project is different. In our free discovery call we can