Update README.md

README.md

@@ -12,12 +12,40 @@ short_description: AI-powered reading coach that adapts to help you learn!
 tag: agent-demo-track
 ---
 
-# ReadRight - AI Reading Coach
+# 🦉 ReadRight - AI Reading Coach for Kids
 
+## Overview
 ReadRight is a simple app designed to help students, especially in non-native English speaking countries, improve their English reading skills. Many students don’t have access to personal tutors or extra help with pronunciation and reading fluency. This app uses AI to fill that gap, making it easier for anyone to practice reading out loud and get helpful feedback.
 
 With ReadRight, students can read short, age-appropriate stories generated just for them. The app listens as they read, checks their accuracy, and gives friendly, easy-to-understand feedback. If a student struggles with certain words, the app will help them practice those words and even create new stories to reinforce learning.
 
-This project is an MVP (Minimum Viable Product) and there is plenty of room for improvement. Features like more languages, better feedback, and a smoother user experience can be added in the future. The goal of ReadRight is to make reading practice accessible, supportive, and fun for everyone, no matter where they live or what resources they have.
+This project is an MVP (Minimum Viable Product) submitted for **Track 3: Agentic Demo Showcase** at the hackathon, showcasing the power of AI agents through a multi-component architecture built with Gradio, smolagents, and advanced AI services.
 
+## 🎯 Motivation
+Non-native English speakers often struggle with pronunciation due to the language’s tricky phonetics (e.g., "knight," "through," "psychology"). Many lack access to affordable, patient, and judgment-free practice environments, leading to embarrassment and reluctance to speak. ReadRight solves this by offering personalized, AI-driven reading practice that adapts to each student’s level and interests, fostering confidence and fluency.
+
+## 🛠️ Technical Architecture
+ReadRight leverages a modular architecture integrated via Gradio Spaces API, utilizing AI agents powered by smolagents in two distinct phases to provide a dynamic, adaptive learning experience. The system employs a multi-step agentic workflow where the LLM’s outputs control the program’s flow, particularly in adapting content based on student performance.
+
+1. **Content and Audio Generation**:
+   - A Gradio interface collects student details (name, grade, topic).
+   - **Story Generation Agent**: Using Google Gemini and smolagents’ tool-calling capabilities, this agent autonomously generates engaging, personalized stories tailored to the student’s grade and interests. The agent dynamically adjusts story length, vocabulary, and complexity based on the student’s grade level, ensuring age-appropriate content without explicit user instruction beyond initial inputs. For example, it selects simpler words for younger students or more complex sentences for older ones, making decisions on content structure internally.
+   - **Audio Synthesis**: Hugging Face TTS (NihalGazi/Text-To-Speech-Unlimited) converts stories into natural-sounding audio for pronunciation guidance. This phase uses LLM outputs as a processor, initiated by user input, but the story generation agent exhibits autonomy in crafting tailored content.
+
+2. **Adaptive Feedback and Learning**:
+   - After the student records their reading, the system activates a multi-step agentic workflow powered by smolagents:
+     - **Speech Recognition**: Whisper Large V2 (abidlabs/whisper-large-v2) transcribes student recordings accurately.
+     - **Text Comparison**: A custom Python engine with `difflib` compares the transcription to the original text, identifying errors and mispronunciations.
+     - **Feedback Generation Agent**: Leveraging smolagents, the LLM generates detailed, encouraging feedback with pronunciation tips, adapting to the student’s performance. It autonomously decides to generate new stories incorporating previously missed or mispronounced words, based on feedback analysis, forming a loop where the LLM determines the next action to address learning gaps.
+   - This phase operates as a multi-step agent, where the LLM controls iteration and program continuation by analyzing performance data and adapting content without explicit user instruction, creating a tailored learning path.
+
+The combination of these phases creates a semi-autonomous system: the content generation phase relies on user-initiated inputs but includes an agentic story generation process, while the adaptive feedback phase demonstrates stronger agentic behavior by dynamically adjusting to student needs, making ReadRight a powerful example of AI-driven educational support powered by smolagents.
+
+## 🎥 Demo
+📺 [Watch the ReadRight Demo Video](#) *(Link to be added)*
+
+---
+
+
+Let’s make reading practice accessible and fun for every student! 🌍📚