Voyager
Guiding Early-Stage Founders with AI-Powered Business Planning
Context
EcosystemOS was preparing to launch Voyager, their AI-powered guidance platform for startup founders, into an increasingly competitive market. As the sole designer assigned to the flagship Business Plan Assistant feature, I was responsible for not only designing this critical product, but also establishing the design language that would unify the entire EcosystemOS ecosystem, including multiple complementary apps being developed in parallel.
Team contex
Solo designer on the Business Plan Assistant, collaborating with:
2 stakeholders (product/business)
2 developers
2 designers working on complementary apps that would adopt this design foundation
Services
UI/UX Design
Component Library & Style Guide
User Research
Wireframing & Prototyping
Landing Page Design
My Role
Associate UX/UI Designer, Design Lead for Business Plan Assistant
My Role
As the sole designer on the Business Plan Assistant, I led the end-to-end design over a 5-month engagement. Working directly with stakeholders and developers, I designed 60+ screens and 100+ reusable components that became the design foundation for the entire platform. My work transformed abstract AI capabilities into an intuitive experience and created the scalable component library that enabled the broader design team to maintain consistency across multiple products.
Impact
Design Foundation for Entire Platform
As sole designer, I established the visual language for EcosystemOS:
60+ reusable Figma components adopted by 2 additional designers across complementary products
Component library continued serving the platform after my engagement ended
Design patterns influenced all subsequent AI features across the ecosystem
Successfully Addressed Core User Needs
Stakeholder validation confirmed the design solved all three key pain points:
Information became significantly easier to digest through progressive disclosure
Founders felt more confident in their decision-making process
AI guidance felt contextual and personalized rather than generic
Independent Execution
Delivered complete feature as solo designer:
Worked autonomously while staying aligned with product vision
Led weekly design reviews with stakeholders and developers
Created landing pages to support product launch
Multiple iteration cycles maintaining quality under aggressive timeline
Understanding the Problem
Early-stage founders face a paradox: access to more resources than ever, yet feeling more lost. Through empathy mapping and stakeholder collaboration, three critical pain points emerged:
Information Overload : Countless frameworks and templates, but no clarity on what applies to their specific situation. Resources became paralyzing rather than helpful.
Lack of Direction :Understanding concepts theoretically but struggling to translate knowledge into concrete next steps. A significant gap between knowing and doing.
Limited Personalization: Existing tools took a one-size-fits-all approach, feeling generic and disconnected from each founder's unique reality.
The guiding question: How might we help founders feel confident and supported without adding to their cognitive burden?
From that starting point, I translated the concept into a clear, personalized experience that adapts to the founder’s stage and needs, turning vague ideas into actionable steps.
Establishing the Foundation
Design Principles
Working as sole designer, I knew my decisions would impact beyond just this feature. With other designers building complementary apps, I established principles to guide the entire platform:
Consistency: Unified components working seamlessly across all features
Hierarchy :Guide attention to what matters most, revealing complexity progressively
Scalability :Design patterns that could adapt and be adopted by other designers
Finding the Right Voice for AI
One of the most significant pivots came from understanding how AI should "speak" to founders. My initial designs approached it like a chatbot, founders would ask questions, AI would respond. It felt transactional, like a support tool rather than a collaborative partner.
The Turning Point
During design reviews with stakeholders, developers, and fellow designers, feedback revealed a critical insight: the AI felt reactive rather than proactive. Founders needed guidance that felt like having an experienced advisor in the room.
I explored multiple approaches:
Voice-Based Interface :Natural but introduced technical complexity and privacy concerns
Traditional Chatbot: Gave control but missed opportunities for proactive guidance
Guided Interview: AI takes the lead, positioning Wisdom as an expert guide
The Solution: Hybrid Approach
We landed on combining the best of both:
AI-initiated guidance as primary flow Wisdom asks strategic questions
Open input option for founders who prefer providing information in their own words
Contextual suggestions appearing as founders write, helping without interrupting
This honored different working styles while maintaining a conversational, supportive tone. The decision influenced how AI operates across all EcosystemOS features and became the foundation for complementary apps.
Collaborative Design Process
Working as sole designer didn't mean working in isolation. I strategically used weekly design reviews to validate ideas and gather diverse perspectives:
Stakeholders provided insight into founder needs and business priorities
Developers offered technical perspective on feasibility and AI limitations
Fellow Designers gave UX feedback and helped ensure my component library would work for their products
This approach allowed me to work independently while staying aligned, moving quickly without working in a vacuum.
From Wireframes to High-Fidelity
After achieving alignment on flows and interaction patterns, I moved into high-fidelity design, developing a comprehensive style guide:
Visual Identity
Color Palette: Cool tones with vibrant blues conveying innovation and reliability
Typography: SF Pro for modern aesthetic and exceptional legibility across devices
Building the Component Library
I built 60+ components knowing they needed to serve multiple purposes: enable my rapid iteration, facilitate developer handoff, and provide foundation for other designers.
Each component included multiple states and variations. Organization and documentation were critical other designers would use these without my direct involvement. I structured the library logically, created clear naming conventions, and documented usage guidelines within Figma.
The investment paid off: when other designers began building complementary apps, they successfully adopted the library, ensuring visual consistency across the entire ecosystem.
Supporting the Launch
As the feature neared completion, I designed landing pages to support product launch, communicating the value of AI-powered business planning while maintaining visual consistency with the platform.
The Result
The final Business Plan Assistant provided founders with a clear, guided path through business planning. By progressively revealing complexity and contextualizing AI insights, we transformed an overwhelming task into a manageable, motivating experience.
Key Features:
Contextual AI Guidance :Personalized insights adapting to each founder's stage and industry
Progressive Disclosure: Complex concepts introduced gradually, preventing cognitive overload
Clear Next Steps: Every AI insight translated into actionable recommendations
Consistent Experience: Component library ensured cohesive interactions
Stakeholder Validation
Stakeholders were particularly impressed with:
My ability to adapt and work independently while staying aligned with product vision
How the component library enabled the broader design team to maintain consistency
The clarity of the final user experience
Production-ready deliverables requiring minimal developer iteration
What I Learned
Leading Design Independently: Working as a sole designer taught me full ownership of design decisions while staying collaborative. Without another designer to validate ideas with, I developed a stronger conviction in design rationale and learned to articulate decisions clearly to non-designers.
Understanding AI to Design Better: I studied ChatGPT and Claude extensively to understand their patterns, limitations, and interaction models. This research directly informed design decisions around displaying AI suggestions, timing loading states, and making AI feel helpful rather than overwhelming.
Designing for Other Designers: Creating components other designers would use required thinking beyond my immediate needs. I had to anticipate use cases across different products, create meaningful variations, and organize everything logically. When other designers successfully adopted the library without significant questions, it validated the approach.
Balancing Speed and Quality: Delivering 60+ screens and 100+ components in 5 months required strategic prioritization. I focused on identifying which details were critical and which could evolve over time, starting with essential variations and adding more as needs emerged.
The Power of Collaborative Independence; Using weekly reviews strategically allowed me to gather valuable input from stakeholders, developers, and fellow designers while maintaining design ownership. This balance of autonomy and collaboration kept me aligned without slowing momentum.
What Happened Next
While my internship concluded after 5 months, the foundation I established continued serving the platform:
Component library became the design foundation for multiple products across EcosystemOS
Two designers building complementary apps successfully adopted the library, ensuring visual consistency
AI interaction patterns I established informed additional features across the platform
The Business Plan Assistant set the quality bar for future work
This project reinforced that great design extends beyond individual screens—it's about creating systems that enable teams to build consistently excellent products.