EVO Prompts

Inspiration

We were inspired by the idea of making educational content more accessible and inclusive for everyone. Many students face unique learning challenges, from dyslexia to visual or auditory impairments that make traditional learning formats difficult to navigate. Our goal was to build a single platform that transforms complex educational materials into formats that are easier to understand, listen to, or interact with, empowering all learners to thrive.

What It Does

STEMVoice is an all-in-one accessibility platform that bridges the gap between educational content and learners with disabilities.

Text-to-Speech: Converts uploaded files, such as PDFs, PNGs, and JPGs into natural-sounding audio using ElevenLabs, allowing users to listen to their materials rather than read them.
Speech-to-Text: Transcribes video or audio uploads (MP3/MP4) into accurate text transcripts powered by ElevenLabs’ speech recognition, making video content accessible to users who are deaf or hard of hearing.
Equation Recognition: Interprets handwritten or printed mathematical equations using MathPix’s OCR API combined with the Gemini 2.5 Flash model, transforming them into readable and analyzable digital text.

Together, these features make STEMVoice a truly multi-modal learning assistant that adapts to the needs of every user.

How We Built It

We began by designing our mockups in Figma Make, focusing on clean, accessibility-driven layouts and responsive design. From there, we developed the web application by integrating:

ElevenLabs API for both text-to-speech and speech-to-text capabilities
MathPix OCR API for parsing equations and handwritten content
Gemini 2.5 Flash model for intelligent summarization

Each feature was carefully integrated and tested to ensure smooth performance and compatibility within the web app’s architecture.

Challenges We Ran Into

One of the biggest challenges was managing API integration between multiple services while maintaining fast response times and consistent output formats. We also had to overcome issues related to file compatibility, ensuring smooth uploads and conversions for a wide range of document and media types. Additionally, fine-tuning accessibility features, particularly accurate speech recognition and mathematical parsing required detailed testing and iteration.

Accomplishments We’re Proud Of

We’re incredibly proud of building a fully functional accessibility platform that leverages ElevenLabs for both speech generation and transcription, alongside MathPix and Gemini for content comprehension and summarization. This was our first time working with these APIs, and seeing them come together seamlessly was a major milestone for our team.

What We Learned

Throughout this project, we learned how to coordinate multiple APIs effectively, handle various file formats, and prioritize accessibility-first design. We also gained valuable experience in team collaboration, version control, and ensuring that our technical implementations truly serve the end user.

What’s Next for EVO Prompts

Moving forward, we plan to expand STEMVoice with real-time collaboration tools and personalized accessibility settings. This includes features such as voice-controlled navigation, adjustable reading speeds, automatic quiz generation from lectures, and AI-powered note summaries.

Our long-term vision is to make STEMVoice a comprehensive accessibility suite that transforms how educational content is experienced by all learners.