AWS Builder Center
Equalize: Breaking Down Language Barriers in Education with AWS

Equalize: Breaking Down Language Barriers in Education with AWS

Every classroom speaks. Not Everyone hears! This is my submission for Bharat Builds 2026 - First Commit. Team ID - NSBJFJ

Software Engineer
Walking into a university lecture and not understanding a single word isn’t a rare occurrence—it’s the daily reality for millions of international and regional students worldwide. While lectures are often delivered in English or a dominant regional language, students whose native language is neither find themselves at a massive disadvantage. The knowledge gap compounds every single day, not because they lack intelligence, but simply because they lack comprehension.
Language should never be the barrier between a student and their education.
That’s why I built Equalize for the Bharat Builds Tour - First Commit Hackathon 2026. Equalize is an iOS app that leverages the power of a fully serverless AWS architecture to instantly transcribe, structure, and translate entire lectures into a student’s native language—complete with AI-dubbed audio, interactive notes, and smart flashcards.
Here is a look at how it works and the architecture that powers it.



🎓 The Equalize Experience

Equalize is designed to be the ultimate companion for any student. You simply open the app, hit record during a lecture (or import an existing audio/video file), and let the AI take over.

📝 Instant AI-Structured Notes

Instead of frantically scribbling during class, students get beautiful, chapter-by-chapter Markdown notes. The AWS cloud pipeline automatically processes the transcript, pulling out key concepts, formatting them into bullet points, and even generating Mermaid diagrams to visualize complex topics.

🌍 One-Tap Language Translation

With a single tap, the entire lecture—notes, transcript, and audio—switches to Hindi, Spanish, or French. The translation pipeline runs server-side and caches locally, meaning students can study entirely in their most comfortable language.

🗣️ Full-Lecture AI Voice Dubbing

Reading translated notes is great, but hearing the lecture is better. Using on-device text-to-speech, Equalize dubs the entire translated lecture transcript aloud in the student’s native language with natural-sounding voices, allowing for an immersive learning experience.

🔍 Select-to-Explain (Apple Intelligence)

Stuck on a specific concept in the notes? Highlight any text and tap the floating "Explain" button. Using on-device Apple Intelligence models, the app generates a friendly, analogy-driven explanation tuned for beginners—powered 100% on-device with zero latency.

🃏 Active Recall & Flashcards

Equalize doesn't just passively give you notes; it helps you study. With one tap, the AWS backend generates a full multiple-choice quiz based specifically on that lecture's content, complete with a detailed performance breakdown at the end.

📴 Offline-First: Record Now, Process Later

University Wi-Fi is notoriously unreliable. Equalize is built to be an offline-first experience. You can record a lecture entirely offline; the app queues it up securely on your device. The moment  NWPathMonitor  detects a restored connection, it automatically fires off the upload and processing jobs in the background without any manual intervention.



☁️ The Engine Room: AWS Serverless Architecture

Equalize relies on a powerful hybrid approach: On-Device Offline capabilities + Cloud-Scale AWS processing. Every heavy computation is offloaded to a fully serverless, auto-scaling AWS backend, while everyday studying happens completely offline. It relies on a seamless hybrid approach: your phone handles the everyday studying, while the AWS Cloud handles the heavy lifting.
I built the app to be "Offline-First." Here is how that looks in practice:
  1. Capture Anywhere (Offline): You can record a lecture even if you have zero bars of Wi-Fi. The app securely saves the audio directly to your device.
  1. The Cloud Engine (Online): The moment you reconnect to the internet, Equalize automatically syncs your recording to our AWS serverless backend. Behind the scenes, a powerful orchestration engine processes the audio, chunks it up, transcribes it, and uses AI to generate your structured notes, flashcards, and translations.
  1. Study Anywhere (Offline Again): Once the cloud is finished, the app downloads everything right back to your phone. From that moment on, you can read your notes, take quizzes, listen to dubbed audio, and ask the AI questions—all completely offline.
How the AWS Components Work Together:
  1. Amazon S3: Securely stores the raw uploaded audio/video files.
  1. AWS Step Functions (The Real Karta Dharta): The true orchestrator of the entire system. It seamlessly coordinates the multi-step pipeline (transcription → chunking → LLM processing) handling fallbacks and state transitions without timeouts.
  1. Amazon Transcribe: Converts the raw lecture audio into a highly accurate, timestamped speech-to-text transcript.
  1. AWS Lambda (Chunker): Intelligently chunks the transcript so it fits perfectly into the LLM context windows.
  1. Amazon Bedrock: Provides the LLM capabilities for generating structured Markdown notes, creating flashcards, and translating content. (Orchestrated alongside other models based on rate limits).
  1. Amazon DynamoDB: An ultra-fast NoSQL database that stores all the processed course content and chapters before the iOS app pulls them down.
  1. Amazon Cognito: Handles secure user authentication and session management.
Once the processing is done, the iOS app fetches the payload and writes it directly to a local SwiftData cache. From that moment on, the student can access everything—notes, transcripts, flashcards, and dubbed audio—completely offline, forever.

Architecture Diagrams

1. High Level Design





End-to-End Workflow



🚀 The Future of Accessible Education

Equalize isn't just a transcription tool; it's a personalized, localized tutor in your pocket. By combining the on-device capabilities of iOS with the massive, scalable power of the AWS cloud ecosystem, we can ensure that every student has an equal opportunity to learn, regardless of the language spoken at the front of the classroom.
Every student deserves to learn in their own language. With Equalize, now they can.
Check out the full  Equalize Video Demo  and if you're interested in checking out the code or trying it yourself, you can find the open-source repository at  Equalize on GitHub .
Any opinions in this article are those of the individual author and may not reflect the opinions of AWS.
Enjoyed reading this content? Let the author know!

Your likes, comments, shares, and saves help creators reach more builders.

Loading recommendations

Loading article