TopicMate — Project Report

Developed by: Varshith Kumar Chinthapatla

Date: September 29, 2025

Version & Purpose

Version: 1.0 — Minimum Viable Product (MVP) + Optional feature roadmap

Purpose: This document describes TopicMate — an AI-powered study companion mobile application
that accepts a student’s topic (typed or spoken), provides concise and accessible summaries, generates
topic-specific practice questions tailored for both competitive exam preparation and slow learners, and
offers supporting features such as progress tracking, flashcards, voice interaction, and gamified daily
challenges.

Table of Contents
1. Executive Summary
2. Problem Statement
3. Solution Overview
4. Target Users & Personas
5. Core Features
6. Optional & Bonus Features (Expanded)
7. User Flows & Wireframes (textual)
8. Technical Architecture
9. API Design & AI Integration
10. Data Model & Database Schemas
11. UX & Accessibility Considerations
12. Security & Privacy
13. Monetization Strategy
14. Deployment Plan & Costs
15. Timeline & Milestones
16. Testing Strategy
17. Analytics & KPIs
18. Marketing & Growth Ideas
19. Appendices (Sample Prompts, Sample Chats, Mock Data)

1. Executive Summary
TopicMate is a mobile-first educational app that leverages modern language models to convert a short
student prompt — a topic — into an immediate learning session. The app will provide:

• A simple, clear summary in two modes (Slow Learner / Competitive).

1
 • Automatically generated practice questions (MCQs, short answers, conceptual) with
explanations.
• Adaptive study plans and progress tracking to personalize learning.

The MVP focuses on robust AI-generated summaries and high-quality question generation, with
optional features (voice input/output, flashcards, visualizers, gamification) slated for subsequent
releases.

2. Problem Statement
Students often face two primary pain points:

1. Time-constrained revision — in competitive exam preparation, students need quick, high-yield

summaries and practice.
2. Accessibility & varied pace — slow learners need simplified, scaffolded explanations and
practice tailored to incremental improvement.

Existing solutions either focus on flashcards (low contextual depth) or full-length lectures (time-
consuming). TopicMate aims to fill the gap with bite-sized, topic-focused AI tutoring.

3. Solution Overview
TopicMate accepts a topic (text or voice), then executes three parallel tasks:

1. Generate dual-mode summary — Fast/Exam-oriented and Simplified/Step-by-step.

2. Generate practice questions — Different formats and difficulties, with model-provided
explanations.
3. Track & adapt — Store responses, score performance, and adapt future questions/study plan.

Optional advanced services include diagrams, flashcards, audio narration, daily challenge quizzes, and
collaborative leaderboards.

4. Target Users & Personas

• Competitive Exam Student (Arjun, 18): Prepares for national entrance exams; needs concise,
high-yield summaries and difficult MCQs; motivated by timed quizzes and leaderboards.

• Slow Learner (Meera, 14): Needs simplified language, step-by-step explanations, hints, and
repeated practice; benefits from audio narration and concept visuals.

• Teacher/Coach (Ms. Rao): Wants to provide targeted practice sets to students and track class
performance.

• Parent (Mr. Sharma): Monitors progress and looks for steady improvement and streaks.

2
5. Core Features (MVP)

5.1 Topic Input

• Text input box with suggestions/auto-complete.

• Optional voice input (speech-to-text) for accessibility.

5.2 Dual-Mode Summary

• Slow Learner Mode: stepwise explanation, definitions, examples, optionally longer with micro-
tutorial steps.
• Competitive Mode: concise summary, key formulas, keywords, and high-yield points.

5.3 Question Generation

• MCQs (4 options), short answer prompts, conceptual/application questions.

• Difficulty tags: Easy / Medium / Advanced.
• Explanations for each answer with references to the summary.

5.4 Chat Interface

• Conversational UI where student interacts with the AI; summary appears as an initial message
followed by question rounds.
• Buttons for Start Quiz , Get More Examples , Create Flashcards .

5.5 Progress Tracking

• Store results per session.

• Weekly summary reports and weak-topic suggestions.

5.6 Export & Share

• Export session summary and questions as PDF.

• Share via link or messaging apps.

6. Optional & Bonus Features (Expanded)

All of these features are included in the product plan and can be prioritized into later releases:

6.1 Voice Input & Voice Output (TTS)

• Speech-to-text to capture topic and short answers.

• Text-to-speech to read summaries, questions, and explanations aloud.

6.2 Flashcards Generator

• One-click create flashcards from summary: term-definition, Q-A flashcards.

• Spaced repetition scheduling (e.g., Leitner system or SM-2 algorithm) integration.

3
6.3 Daily AI Challenge

• A short daily quiz (3–7 questions) personalized by recent performance.

• Streaks, XP, and badges to encourage consistency.

6.4 Leaderboards & Social Features

• Weekly global/class leaderboards.

• Option to create private groups (class/college) and challenge peers.

6.5 Study Plan Generator

• AI builds a weekly plan based on syllabus length and student performance; includes daily goals
and topic recommendations.

6.6 Concept Visualizer & AI Diagrams

• Generate simple diagrams or mind maps from the summary using a diagram generator (vector
PNG/SVG).
• Useful for biology processes, mathematical graphs, historical timelines.

6.7 Notes Export & Sharing (Advanced)

• Export as PDF, DOCX or Markdown.

• Teacher export mode for batch printing of practice sets.

6.8 Accessibility Features

• Adjustable text size, dyslexia-friendly fonts, color contrast options, and read-aloud controls.

6.9 Offline Mode (Cached Content)

• Cache recent summaries, flashcards, and question sets for offline revision.

6.10 Multilingual Support

• Summaries and questions can be generated in multiple languages (e.g., English, Hindi, Telugu)
using language setting and translation pipelines.

6.11 API for Schools & Integrations

• Provide an API for LMS (Learning Management Systems) to pull content, or for teachers to push
topic sets to students.

6.12 Monetization Add-ons

• Premium subscription for advanced features (detailed study plans, advanced mock tests, offline
packs, exported PDFs).
• In-app purchase for mock test packs and topic bundles.

4
7. User Flows & Wireframes (Textual)

7.1 New User Onboarding

1. Welcome screen → Signup (email/phone/Google) → Quick skill-level quiz (optional) → Home feed
with suggestions.

7.2 Single-Topic Session

1. Student enters topic (or speaks).

2. AI returns dual-mode summary.
3. Student selects mode and difficulty.
4. AI generates 5–10 questions; student answers in chat.
5. AI provides explanations and updates progress.

7.3 Flashcard Session

1. From summary, student taps Create Flashcards .

2. Student selects which cards to save.
3. Repetition schedule is created and reminders set.

7.4 Daily Challenge Flow

1. App pushes daily challenge notification.

2. Student completes timed quiz.
3. Score is added to leaderboard and progress is updated.

8. Technical Architecture

8.1 High-Level Components

• Mobile Client (React Native) — Chat UI, local storage, audio I/O.
• Backend API ([Link]/Express) — Auth, session management, caching of generated content,
analytics.
• AI Layer (OpenAI or equivalent) — Summary generation, question generation, difficulty
control, explanation generation.
• Database (Firebase / MongoDB) — Users, sessions, progress, flashcards, leaderboards.
• Storage (Cloud Storage) — Exported PDFs, generated images/diagrams.

8.2 Sequence: Topic Session

1. Client sends POST /session with topic + user preferences.

2. Backend validates, normalizes topic, and checks cache.
3. If cached, return cached summary+Q set. Else:
4. Call AI to generate summary (two templates: Slow, Competitive).
5. Call AI to generate question set (tag each question with difficulty and explanation).
6. Store results in DB and return to client.

8.3 Caching Strategy

• Keep a short-term cache for high-frequency topics to reduce AI API calls.

5
 • Respect user privacy — cached items are associated with user id and optionally anonymized.

9. API Design & AI Integration

9.1 Example Endpoints

• POST /api/v1/session — start session (body: topic, mode, difficulty).

• GET /api/v1/session/:id — retrieve session results.
• POST /api/v1/answer — submit answer for scoring and feedback.
• POST /api/v1/flashcards — create flashcards from session data.

9.2 AI Prompting Strategy (Examples)

• Summary prompt template (Slow Learner):

You are an experienced teacher explaining to a young student. Summarize

the topic "{topic}" in simple step-by-step language, include
definitions, examples, and 3 small practice prompts with hints. Keep
each sentence short and use simple vocabulary.

• Summary prompt template (Competitive):

You are an expert tutor for competitive exams. Provide a concise, high-
yield summary for "{topic}" including key formulas, important facts, and
a one-line memory tip. Keep it under 200 words.

• Question generation prompt:

Generate {n} questions on "{topic}" across difficulty levels [Easy,

Medium, Advanced]. Provide for each question: type (MCQ/Short/Concept),
the correct answer, three distractors for MCQ, and a brief explanation
tying back to the summary.

9.3 Safety & Hallucination Mitigation

• Use structured prompts asking the model to cite (if possible) or to indicate confidence levels.
• For critical or factual topics, cross-check with a small curated knowledge base when feasible.

6
10. Data Model & Database Schemas (High-Level)

10.1 User

User {
id: UUID,
name: string,
email: string,
preferences: { language, mode, difficulty },
createdAt, updatedAt
}

10.2 Session

Session {
id: UUID,
userId: UUID,
topic: string,
mode: enum(Slow, Competitive),
summary: { slowText, competitiveText },
questions: [ { id, type, difficulty, questionText, options, correct,
explanation } ],
score, createdAt
}

10.3 Flashcard

Flashcard {
id, userId, front, back, interval, nextReviewDate
}

10.4 LeaderboardEntry

{ userId, score, date, scope (global/class) }

11. UX & Accessibility Considerations

• High-contrast themes and adjustable text sizes.
• Dyslexia-friendly fonts.
• Clear microcopy and hint buttons for each question.
• Voice interaction to support non-typists and visually impaired learners.

7
12. Security & Privacy
• Authentication: Firebase Auth / OAuth2.
• Data protection: Encrypt sensitive fields at rest.
• User control: Users can delete stored sessions and export or remove all data.
• GDPR/Local compliance: Add terms for data retention, export, and deletion.

13. Monetization Strategy

• Freemium model: Basic summaries and limited questions free.
• Subscription tiers: Monthly / Yearly with features: unlimited sessions, advanced mock tests,
offline packs, multi-language packs.
• One-time purchases: Topic bundles, mock test packs.
• Institutional licensing: Schools and coaching centers with bulk pricing and teacher
dashboards.

14. Deployment Plan & Cost Estimate (Initial)

14.1 Infrastructure

• Backend: Small [Link] server (hosted on Vercel / Render / DigitalOcean).

• Database: Firebase (starts free, scales) or MongoDB Atlas.
• AI API: OpenAI usage costs proportional to tokens; set budgets and caching to control costs.

14.2 Estimated Monthly Costs (MVP scale)

• Hosting & DB: $20–$100

• Storage: $5–$50
• AI API usage: $100–$1000+ (highly variable—monitor usage)
• Analytics & Push notifications: $10–$50 n

Note: These are rough estimates. Actual costs depend on active user base and
token usage per session.

15. Timeline & Milestones (12-week expanded plan)

• Weeks 1–2: Design & prototyping (UI flows, wireframes).
• Weeks 3–4: Core chat UI + Backend + Auth.
• Weeks 5–6: AI integration (summary + question generation), caching.
• Weeks 7–8: Progress tracking, flashcard generation, export functionality.
• Weeks 9–10: Voice features, diagrams, offline caching.
• Weeks 11–12: Testing, polishing, beta release, marketing prep.

8
16. Testing Strategy
• Unit tests for backend logic and question scoring.
• Integration tests for API + AI prompt responses and validity checks.
• User testing with both competitive-prep students and slow learners to evaluate clarity and
helpfulness.
• A/B testing for summary length and question difficulty tuning.

17. Analytics & KPIs

• Daily Active Users (DAU), Monthly Active Users (MAU)
• Average session length (time per topic)
• Accuracy rates per topic and difficulty
• Retention / churn rates
• Conversion rate from free to paid

18. Marketing & Growth Ideas

• Partner with coaching centers and schools.
• Create short social media videos with daily challenge highlights.
• Offer teacher dashboards and pilot programs for class adoption.
• Referral program (invite friends → unlock premium days).

19. Appendices

Appendix A — Sample Prompts (for AI)

Slow Learner summary prompt:

You are a patient teacher. Explain the topic "{topic}" at a middle-school

reading level. Use short sentences, numbered steps, simple examples, and
include 2 small practice prompts with hints.

Competitive summary prompt:

You are a concise exam coach. Summarize "{topic}" in high-yield format: key
points, formulas, and a 1-line memory trick. Keep under 200 words.

Question generation prompt:

9
 Generate 8 questions on "{topic}": 3 Easy MCQs, 3 Medium MCQs, 2 Advanced
conceptual questions. For each MCQ provide 3 distractors, indicate the
correct option, and write a 1–2 sentence explanation referencing the summary.

Appendix B — Sample Chat Transcript

Student: "Today's topic: Photosynthesis"

AI (Slow Learner): "Photosynthesis is the process by which plants make food using sunlight. Step
1..." (short stepwise text)

AI (Quiz): "Q1: Which gas is produced? A) CO2 B) O2 C) N2 D) H2 — Correct: B; Explanation..."

Appendix C — Mock Data Example (Session)

Session {
id: s1,
userId: u1,
topic: "Photosynthesis",
mode: "Slow",
summary: { slowText: "..." },
questions: [ { id: q1, type: MCQ, difficulty: Easy, questionText: "...",
options: [...], correct: "B", explanation: "..." } ],
score: 4/5
}

Closing Notes & Next Steps

I have included all core, optional, and bonus features discussed. Next options you may want me to
perform now:

• Generate a PDF export of this report.

• Create AI-generated prototype images (diagrams, app mockups) to include in the PDF.
• Produce a React Native starter template (single-file sample) with the chat UI and AI integration
hooks.

Please tell me which of the above you'd like next — I can generate the PDF and add prototype images
right away.

10