Client

Student-friendly LLM built with OpenAI and Anthropic models

Viacela is an AI-powered platform designed to amplify students' authentic voices. By fostering frictionless reflection, we create dynamic profiles that capture both tangible achievements and intangible growth, empowering educators to personalize support and guiding students toward career readiness.

I was brought on as the founding product designer and leading design voice with Richard __ (prev CTO of Command Bar) and George Tang to co-create the Viacela MVP for beta launch in January.

Problem

On average, students receive only 40min of career counseling per year

Educators are unable to deeply understand a student's strengths, developmental areas, and aspirations. Traditional methods fail to provide personalized, actionable insights, leaving students without the tailored support needed to navigate individual pathways and achieve career readiness.

Schools lack the staff and resources to ensure every kid's needs are visible and supported.

My role

Evolving a demo app to MVP for live user testing

As the Founding Product Designer at Viacela, I worked closely with the CEO and founding engineer to define the MVP's goals and scope. In a highly ambiguous environment, I led structured brainstorming sessions to align the team's vision and clarify objectives. I introduced structure to our workflows by creating internal documentation and SOPs. Collaborating with the founding engineer, I ensured designs balanced innovation with technical feasibility, leveraging a deep understanding of LLM capabilities. I also drove ongoing discussions to prioritize features that would balance client needs with student engagement, aiming for product-market fit.

Current demo

Business requirements

Designing responsive experiences for 3 users: students, teachers, and parents

Define UI/UX experience and create high fidelity and responsive designs, optimized for mobile and tablet use:

Student interface

• Daily reflections

• Viewing past entries

• Viewing the latest AI-generated profile

Admin & teacher interface

• Viewing student data, surveys, and profiles

• Validating and confirming upcoming AI-generated survey questions

• Validating and confirming upcoming AI-generated student profiles

• Suggesting questions/edits

Parent interface

• Viewing student profile

• Suggesting questions/edits

📱 Final MVP - Mobile web

Student flows

Admin & teacher flows

💻 Final MVP - Web

A peak at the web experience

Challenges

Training the LLM / navigating ambiguity

While high level objectives were clear,

• had to be careful of latency and load times

• due to cost of refreshing profile, we had to set restraints to

• Adapting new design trends

Research

User research on students and schools to understand what kind of information would be helpful for them to help support their students.

Wireframes

Designing for mobile optimization

One key decision I made was how to design the mobile version of the application. Through user research, I discovered a few key things:

1. Users no longer rely on the header

2. Brand identity has changed.

A challenge is designing the app to operate on familiar mental models, yet still step out of the bounds of a normal website, such as an ecommerce website or apps with bottom navigation tabs.

✦✦ The process ✦✦

Want to learn more? Let's chat!

I'm happy to walk through details of my design process through a discovery call. You can reach me at renali.2938@gmail.com or schedule a time on my Calendly.