Thryve

Web design UI/UX

February, 2026

Role/Services

Web design UI/UX

Client

Thryve, Worked at Geekbears

Category & Year

Project Overview

Middle and high school students in Florida spend hours doing homework and test prep — yet careless mistakes, test anxiety, and surface-level memorization remain the norm. The root problem isn't effort. It's that students don't have the right tools to slow down, understand the reasoning behind a problem, and build the mental frameworks needed to solve similar ones on their own.

Thryve was designed as an AI-powered, web-based tutoring platform that transforms passive help-seeking into an active, guided discovery process. Rather than giving students the answer, Thryve's AI tutor detects how each student learns best — starting from an onboarding calibration session — and uses that profile to guide them step by step through problems using a visual whiteboard, AI-generated images, and adaptive analogies. The goal is never to solve the problem for the student. It's to build the understanding that lets them solve the next one.

The platform is fully aligned with Florida's B.E.S.T. Standards for Mathematics and built to support students preparing for FAST Assessments, making it uniquely positioned for the Florida K–12 ecosystem.

The challenge

The core challenge was not just building an AI tutor — it was designing a product that makes guided reasoning feel natural, calm, and accessible for teenagers who are already frustrated with school, without ever handing them the answer.

Key questions

How might we guide students to the answer without ever giving it to them?
How might we detect a student's learning style and adapt explanations in real time?
How might we make structured reasoning feel natural and low-pressure for teenagers?
How might we give parents meaningful visibility into whether real understanding is being built?
How might we design an experience fully aligned to Florida's B.E.S.T. Standards and FAST exams?

Discovery phase

The project began with a strategic discovery process to better understand the ecosystem, user pain points, and market opportunities before moving into design execution.

Discovery methods

Stakeholder discovery questions
User persona development
Competitive UI/UX benchmarking
Feature prioritization for MVP
User journey mapping
Sitemap / information architecture

Understanding the users

Primary user — student Middle and high school students needed support managing:

Students needed support with:

Careless mistakes driven by rushing and lack of structured reasoning
Anxiety and avoidance when concepts don't click in class
Surface-level memorization that falls apart under FAST exam pressure
No feedback loop to understand where their thinking actually breaks down

Secondary user — parent

Parents wanted visibility into their child's learning without micromanaging. They needed to trust that the platform was building real understanding — not just keeping their kid busy — and that progress was tied to something concrete like Florida's B.E.S.T. benchmarks and the FAST exam countdown.

Key insight

The student is never alone in the learning problem — parents carry anxiety, coaches carry pressure, and the platform must serve the whole ecosystem..

Competitive benchmarking

An analysis of existing platforms revealed that most competitors focused heavily on either:

Answer-first

Deliver the solution - Platforms like Photomath and Mathway optimize for getting the correct answer instantly, with no reasoning development or curriculum alignment.

Content-first

Broad AI tutoring - Tools like Studdy AI offer interactive whiteboard sessions across all subjects but without a learning profile, state standards alignment, or parent visibility.

Standards-adjacent

Drill and practice - Platforms like IXL align to state curricula but deliver content through repetitive exercises — not through guided reasoning or adaptive pedagogy.

Market gap identified

There was an opportunity to create a Florida-specific AI tutoring product that balanced:

Guided reasoning over answer delivery
Adaptive learning profiles built from session data
Full Florida B.E.S.T. and FAST alignment
Parent visibility without surveillance

Strategic opportunity

Position Thryve as the first AI tutoring platform built specifically for Florida students — one that teaches students how to think through problems, not just complete them.

Defining the MVP

To avoid feature overload, the product was intentionally scoped around core behavioral outcomes first.

Must-have features

Student onboarding + learning profile calibration
Practice mode (Thryve-driven, B.E.S.T. aligned)
Homework help mode (student-initiated)
Test prep mode (FAST-aligned)
AI whiteboard with step-by-step guided reasoning
AI-generated visual resources per learning profile
Hint system (limited, preserves discovery)
Mistake diagnosis and reasoning feedback
Thryve Index™ + Thinking Score
Parent dashboard with benchmark progress

Deprioritized for later phases

Teacher / coach dashboards
Multi-subject support beyond math
Advanced AI personalization (deep adaptive pacing)
Wearable / LMS integrations
Social and gamification layers
Mobile native app

Product strategy

Focus first on establishing the guided reasoning loop and validating that students genuinely understand concepts — not just complete sessions — before expanding into ecosystem complexity.

User journey

TThe user journey centered on cognitive and emotional progression — building genuine understanding, not just session completion.

Before

Confusion, rushing, surface-level memorization

During

Personalized learning profile, step-by-step guided discovery, adaptive visual reasoning

After

Real conceptual understanding, reduced test anxiety, visible benchmark progress

Core UX principle

Design for understanding, not completion

Information architecture

A simplified sitemap was developed to ensure the product remained approachable for younger users while maintaining personalization.

Student platform — 13 screens

Splash / Auth
Onboarding
Home
Practice session
Homework help
Test prep mode
Settings

Parent platform — 5 screens

Splash / Auth
Parent dashboard
Progress overview
Session overview
Settings

Design priority

Reduce friction into the learning loop while maintaining parental oversight that informs without overwhelming.

Product vision

Thryve was designed to become more than an AI tutoring tool — it aimed to function as a personalized reasoning development system that grows with the student across their entire Florida K–12 journey.

Intended value

Help students build the mental frameworks to solve problems independently — not just complete homework sessions.

Outcome

The discovery and UX strategy phase established a complete product foundation. The project is currently advancing through the remaining phases of execution — from wireframes into high-fidelity UI and brand identity.

Wireframes

A complete wireframe system was built across both platforms — student and parent — covering all MVP screens with annotated dev notes documenting key interaction decisions directly in Figma.

High-fidelity preview

A first pass at the visual design system explores how the wireframe foundation translates into a calm, focused product experience — built around Thryve's teal and coral palette, Nunito typography, and a whiteboard-centered layout designed to reduce cognitive load during active learning sessions.

Logo proposal

The Thryve logomark was designed around two core ideas: the interlocking speech bubbles represent the back-and-forth dialogue between the student and Thryveron, the AI tutor — a conversation that never gives the answer, but always guides toward it. The coral accent dot with radiating marks is the "aha moment," the instant a student stops being confused and starts understanding. Together with the name — a deliberate respelling of "thrive" — the mark signals growth, forward motion, and the idea that real academic confidence is something you build, not something you're given.

Let's Build Something

Extraordinary Get In Touch!