Decode
Morning
Tradl AI
AI-first financial analyst for real-time stock screening, research, and market analysis
My Role:
Product Designer (Intern)
Team:
Senior Designer
A Developer
Time:
20 days
Shipped
Scope:
Web App
Why does product exist?
Tradl built Morning Decode to turn information overload into a clear market view.
Problem
Product was built entirely using AI
View Linkedin Post
To validate market fit rapidly, the founding team launched a zero-code MVP powered by AI. While it proved core demand, user engagement dropped due to high cognitive load and mobile usability friction.
Hurdles found after product evaluation:
Desktop first design, not optimised for mobile view
Multiple CTA for the same flow creates navigation ambiguity.
Users struggle to distinguish between clickable and non-clickable elements.
Users accessed the tool on mobile browsers where touch targets overlapped
Events showed drop-off before reaching section 3.
Research
Data validation became necessary for identified the problems
Even though we were quite stoked about all the ideas, it was crucial that the users also felt the same. Identified the trade-offs and retention rate using software data. Additionally, I took internal feedbacks to gather what team says about the Morning decode.
Research identified 3 high-severity issues prioritized for redesign
Desktop first design, not optimised for mobile view
Team was unsure which action to use when multiple options lead to the same flow.
Information overload with less hierarchy in content placement
Key Design Challenges
01
How might we improve content density and navigation so users can scan daily insights without excessive scrolling?
02
How might we show essential morning decode features without cluttering the core analytical flow?
This started the loop of iterations, testing, interations for more than 12 days to bring the design that solve the problem in more impactful ways.
Testing helped in improving concepts
Below I showed how I improve the concept for making navigation with each testing.
Concept 1
Reduced click depth but failed discoverability tests.
Concept 2
Bottom tabs increased glanceability but made hard to categorise sections.
Concept 3
Combined both style with a promotional button to increase growth.
Designed 10+ iterations in 4 days
Since the requirements of the product was quick I ideated and prototypied simulatenously, which came as impactful decision.
I used Claude Design for faster iterations, this proccess felt like a design sprint.
Presenting
A more impactful morning decode
Web App
A single page to scan the overall market within 30 seconds
A non-distracting promotional button
Fixed space for Subscribe button in nav bar.
Listen to the market summary anytime
No need for long reads, quick 4-5 mins audio episode.
Easier Navigation between the sections
Users can reach any section of the page in just two taps.
Drawers for frictionless user actions
Help in quick sharing, and form entries without losing the context.
From AI to Human Centric Design
Before
After
Variations in the design occurred for two main reasons: multiple developers contributed to the interface, and a unified design system had not been established.
Validation & Success Criteria
Because this design shipped at the conclusion of my internship, long-term retention data was tracked post-handoff. We defined three core signals to evaluate the new design:
Drop-off Reduction
Monitoring scroll depth on mobile to verify if users now reach sections 4–8 without
Scan Speed (Time-to-Insight)
Tracking session duration during market open hours (8:30–9:15 AM) to confirm users can consume data in under 30sec.
Audio Engagement
Tracking play-rate and completion-rate on the 4-minute market summary audio widget.
Learnings
Navigating these constraints provided valuable lessons in system architecture, cross-functional alignment, and designing for real-world user behavior.
Goal-Driven Iteration
Thoughtful content and design decisions reduce unnecessary API calls and overall AI costs.
Goal-Driven Iteration
Building with AI is faster when instructions focus on goals and continuous feedback rather than one-time prompts.
Information Hierarchy
Designing for finance product means organizing data so users can quickly understand it and act, without removing important information.
Thank you for scrolling