Jose AI
AI Operations App
CLIENT
SCALEDai
UI/UX Designer | Team
2 weeks
Tablet
THE PROBLEM
Small business owners struggled to understand AI-driven insights due to complex language, rigid dashboards, and low adaptability to their context.
THE SOLUTION
I designed a touch-first tablet app that translated AI insights into plain-language, adaptive guidance tailored to real small-business workflows.
Clarity Snapshot
An MVP AI business assistant that transforms raw research and small-business data into clear, actionable guidance.
The Chaos
The founder had done extensive research into the pain points of small businesses:
1) lack of time,
2) limited digital literacy, and
3) difficulty interpreting data.
But they didn’t know how to translate those findings into a usable product.
1) lack of time,
2) limited digital literacy, and
3) difficulty interpreting data.
But they didn’t know how to translate those findings into a usable product.
Small business owners were overwhelmed by data and lacked a clear next step, despite having access to “insights.”
The Constraints
2-week MVP timeline tied to investor funding
No opportunity to validate long-term AI behavior
Non-technical, immigrant-owned small businesses as primary users
High risk of mistrust or misinterpretation of AI outputs
Team alignment required across design, marketing, and business goals
The Insight
Testing and user interviews revealed that most small business owners didn’t need more analytics, but needed clear guidance they could trust.
We intentionally designed José at a 13th-grade reading level not to “dumb down” insights, but to meet users where they were emotionally and cognitively.
Many users understood their businesses deeply, but not the language of analytics. The design needed to adapt to how people think under stress, not just what they technically understand.
Many users understood their businesses deeply, but not the language of analytics. The design needed to adapt to how people think under stress, not just what they technically understand.
The Structure
1. Simplicity but not limited
- Every interface is stripped to essentials while still offering access to deeper tools when needed.
2. Guided, easy-to-understand, interaction
- The AI guides users step-by-step, using conversational flows and contextual cues instead of nested menus.
.webp)
3. Modern recognition
- The visual language aligns with modern AI products to immediately signal credibility, clarity, and trust.
.webp)
The Clarity
The resulting MVP translates complex research into a calm, human experience: a platform that speaks to busy business owners in plain language and helps them make smarter decisions, faster.
Onboarding: Removing Friction Early
Design Decision
Progressive commitment over upfront complexity
The onboarding intentionally sequences value-first connection (POS sync) before deeper configuration, reducing early cognitive load while still collecting only the minimum information needed to personalize insights and deliver immediate business value.
The onboarding intentionally sequences value-first connection (POS sync) before deeper configuration, reducing early cognitive load while still collecting only the minimum information needed to personalize insights and deliver immediate business value.
Conversation-First Insights
Design Decision
One key tradeoff was choosing guided conversational flows over traditional dashboards. While dashboards offer flexibility, early research showed they increased cognitive load for our users. We prioritized clarity and momentum over power-user control, knowing this could evolve later.
From Recommendation → Decision
Design Decision
Designed recommendations to be directly actionable by pairing AI insights with immediate financial impact and inline rationale, so users can move from insight to decision without leaving context.
.webp)
What We Intentionally Deferred
- Long-term AI memory and personalization
- Automated decision-making
- Predictive insights beyond near-term recommendations
Given the MVP scope and investor timeline, we focused on building trust first. Adaptability was designed as a future layer, not a launch requirement.
The Impact
Reduced cognitive load by designing interactions that communicated intent through structure and affordance, not instructions
Established a clear interaction model that made complex system behavior feel intuitive without onboarding overhead
Aligned design execution closely with engineering constraints, accelerating implementation readiness
Demonstrated the product’s value proposition in a 1-minute teaser video, translating complexity into clear user outcomes
The Reflection
This project reinforced my approach to product design under constraint: prioritize trust before intelligence, guidance before flexibility, and structure before scale.
When working with AI, the challenge is not making systems smarter, but making decisions feel safe, understandable, and human.
When working with AI, the challenge is not making systems smarter, but making decisions feel safe, understandable, and human.
