LiteraSeed
LiteraSeed is a startup dedicated to advancing health equity for the most vulnerable populations by improving communication between patients and doctors. Recognizing that miscommunication contributes to 78.9% of misdiagnoses and results in significant healthcare costs, LiteraSeed aims to address these challenges through innovative solutions.
Client
LiteraSeed
Deliverables
Visual design Digital Asset Management Web design Workflow
Year
2023 - 2025
Role
Design Lead
One of the primary challenges LiteraSeed addresses is the difficulty many immigrants and refugees face when navigating complex medical forms due to language and literacy barriers. As a designer, this challenge pushed me to think beyond conventional form layouts and explore more inclusive interaction patterns. The team developed a web application designed with visual and auditory support to help patients navigate care in their own language, enabling them to communicate symptoms effectively before seeing a provider. My design choices prioritized clarity, empathy, and universal accessibility— ensuring that patients, regardless of literacy level, could engage with the tool confidently. The application is also HIPAA-compliant, supporting secure communication and timely care delivery.
Patient Questionnaire App
Challenges & Objectives
This project needed to address serious communication barriers in healthcare while remaining flexible enough for broader clinical use.
📌 Challenges:
Many immigrants and refugees struggle with complex medical forms due to language and literacy barriers.
Different healthcare settings—like Emergency, Pediatrics, or OB/GYN—require different types of questions, each with unique needs and constraints that add complexity to the design process.
📌 Objectives:
Create a simple, intuitive, and accessible questionnaire app to improve patient communication before they see a provider.
Build a solid, flexible framework that works well for the initial deployment in the Emergency Department.
Ensure adaptability for use in other departments and future expansions.
Solutions & Implementation
Designed a web-based questionnaire tool enhanced with visuals and audio in patients’ native languages to reduce literacy barriers.
Designed a modular structure that supports multiple question types, including symptom checklists, follow-up logic, open-ended inputs, anatomical selectors, and numeric inputs.
Built in multilingual support and accessibility features like text-to-speech and simplified navigation.
Iterated on the design based on feedback from clinicians and observations from user testing sessions to refine interactions, visual hierarchy, and overall usability.
Impact & Results
Since joining the team, I’ve been leading key design improvements and infrastructure work to scale and extend the impact demonstrated in Phase 1. As of May 2025, the project is scheduled for a clinical trial with Valleywise Health's Emergency Department in Arizona to further validate and expand its effectiveness in real-world settings.
Anatomical Selection Experience
While many question types in the app had clear design solutions, one posed a unique challenge: “Where on your body do you feel pain the most?” For patients with limited language proficiency, selecting a specific body region needed to be intuitive and error-resistant.
Design Challenges
Due to resource constraints, we worked with a 2D illustration model rather than building a full 3D version. Several design goals emerged:
Ensure medically relevant and comprehensive body part selection.
Create a clear visual hierarchy to minimize patient error.
Design natural, culturally sensitive interactions.
Iteration Process
The interaction design evolved through three major iterations:
Initial Approach
The first version used the existing 2D model with connectors between body regions and corresponding answer choices. While this did provide visual indication, it wasn’t always clear—particularly in areas where multiple regions overlapped. As a result, some patients had difficulty recognizing which region was selected or how to correct it.Enhanced Interaction
To address this, we added interactive highlighting. When a patient selected a region (either via button or tapping the model), the corresponding area on the model lit up. This added clarity but still left room for confusion due to indistinct visual boundaries.Direct Selection Model
In the final version, all selectable body regions were made visually distinct from the start. Each region was redrawn with minimum overlaps as an individual vector shape, allowing for clear, direct selection on the model.
Design Handoff & Implementation Prep
The final step was preparing the assets and metadata for handoff to development. The goal was to make everything easy to iterate on, digestible for developers, and ready for implementation.
To keep things efficient, I designed a system using four gendered base models and two non-gendered ones. These served as flexible templates that could generate all the necessary combinations with just a few custom overrides.
Each selectable body region was assigned a unique ID, allowing developers to easily trigger state changes when a patient made a selection. To further support implementation, I created a spreadsheet that included:
A list of all regions and their corresponding IDs
Metadata and visual references
Step-by-step instructions for building the interactions
This setup minimized ambiguity and reduced back-and-forth, ensuring a smooth handoff. Close collaboration with the development team helped preserve design intentions even under tight resource constraints.
Digital Asset Management Workflow
As the scope of the project expanded, managing visual assets became a critical part of ensuring both quality and efficiency. For the Emergency Department questionnaire alone, we needed over 1,000 visual assets.
Challenges
Disorganized and inconsistent files: Source and working files (mostly .ai) were scattered across folders, often poorly organized. Inconsistent file naming, mixed background treatments, and duplicate assets made it hard to find, edit, and manage assets reliably.
Lack of visual consistency: Exported assets varied in dimension, resolution, aspect ratio, and framing—resulting in UI inconsistencies and a product that looked unfinished.
No version control: Without a clear versioning system, multiple versions of the same asset were sometimes used at the same time, leading to confusion and visual mismatches.
All of this led to wasted time, miscommunication, and a product that felt visually inconsistent.
Solution & Approach
We implemented a three-part approach to improve asset management and delivery:
Organization & Cleanup: Audited existing assets to remove redundancies, fixed organization issues, and set up a standardized file structure with consistent naming conventions.
Standardization & Optimization: Created clear export rules for dimension, resolution, background treatment, and visual alignment. Optimized over 1,000 visual assets to be export-ready and visually consistent.
Workflow & Handoff: Documented a scalable workflow, shared the process with the team, and ensured smooth handoff for deployment—making future updates more efficient as the product expands to support other clinical areas.
The new system helped the team quickly find, update, and use assets—making the product feel more consistent, polished, and production-ready.
Conclusion & Key Takeaways
Working on LiteraSeed has been an eye-opening experience—both in the complexity of designing for underserved patients and in the rewarding challenge of contributing to a mission-driven startup. Collaborating directly with the two co-founders gave me rare insight into how vision, grit, and design come together to push healthcare forward.
In a fast-paced, scrappy startup environment, I wore many hats: designing user flows, refining production processes, and even setting up Asana boards to improve team collaboration. Every task, big or small, had a clear connection to patient experience—and that kept the work grounded and meaningful.
This project deepened my belief that thoughtful design isn't just about beauty or polish—it's about enabling access, clarity, and dignity. It reminded me that even the smallest interactions, when designed with care, can have a lasting impact on people’s health and trust.
