VoxScholar.com

Speak your questions. Get patient answers. Learn at your true pace.

Score: 7.9/10BFHard BuildReady to Spawn
Brand Colors

The Opportunity

Problem

Students lack personalized instruction and a judgment-free space to learn at their own pace, causing them to fall behind in one-size-fits-all classrooms.

Solution

VoxScholar is a mobile app delivering personalized instruction through natural voice conversations in English and local languages. Students speak questions and receive immediate, judgment-free explanations, practice, and feedback. The AI continuously adapts difficulty and pacing based on real-time comprehension, creating a completely private learning environment for students who have fallen behind.

Target Audience

K-12 students in under-resourced schools and developing countries (esp. Sub-Saharan Africa and Southeast Asia)

Differentiator

Voice-first design built for oral learning cultures and low-literacy environments common across Sub-Saharan Africa, with strong offline capabilities and regional language support (Swahili, Hausa, Bahasa Indonesia).

Brand Voice

supportive, encouraging, patient

Features

Voice Conversation Interface

must-have65h

Real-time speech-to-text input and natural text-to-speech output for hands-free tutoring

AI Tutor Engine

must-have85h

LLM-powered tutor that generates explanations, examples, and follow-up questions

Adaptive Learning Paths

must-have55h

Automatically adjusts topic difficulty and sequence based on student mastery

Offline Mode

must-have50h

Downloads lessons, models, and exercises for learning without internet

Private Progress Tracker

must-have35h

Visual mastery dashboard only visible to the student

Regional Language Support

must-have50h

Swahili, French, Bahasa, and English with accent adaptation

Anonymous Peer Q&A

nice-to-have40h

Students can post voice questions answered by peers anonymously

Guardian Summary Reports

nice-to-have30h

Opt-in weekly voice or text summaries for parents

Daily Streaks & Rewards

nice-to-have25h

Gentle gamification to build consistent learning habits

Total Build Time: 435 hours

Database Schema

users

ColumnTypeNullable
iduuidNo
phone_numbertextNo
grade_levelintNo
preferred_languagetextNo
created_attimestampNo
onboarding_completedboolNo

Relationships:

  • references subscriptions(id)
  • references progress_records(user_id)

learning_sessions

ColumnTypeNullable
iduuidNo
user_iduuidNo
subjecttextNo
duration_minutesintYes
started_attimestampNo
ended_attimestampYes

Relationships:

  • foreign key user_id to users(id)
  • has many interactions

progress_records

ColumnTypeNullable
iduuidNo
user_iduuidNo
subjecttextNo
mastery_levelintNo
last_updatedtimestampNo
topics_masteredtextYes

Relationships:

  • foreign key user_id to users(id)

subscriptions

ColumnTypeNullable
iduuidNo
user_iduuidNo
tiertextNo
statustextNo
current_period_endtimestampNo
flutterwave_subscription_idtextYes

Relationships:

  • foreign key user_id to users(id)

API Endpoints

POST
/api/sessions/start

Creates new learning session and returns context for AI

🔒 Auth Required
POST
/api/chat/process

Accepts transcribed voice input and returns AI tutor response

🔒 Auth Required
GET
/api/progress/fetch

Returns student's mastery levels and next recommendations

🔒 Auth Required
GET
/api/offline/bundle

Downloads subject bundle for offline use

🔒 Auth Required
POST
/api/webhook/flutterwave

Handles subscription payment events

Tech Stack

Frontend
Flutter (speech_to_text, flutter_tts, riverpod, dio)
Backend
FastAPI (Python)
Database
PostgreSQL
Auth
Supabase Auth (phone-first)
Payments
Flutterwave
Hosting
Railway (API), Supabase (DB/Auth), Play Store + App Store
Additional Tools
Groq + Llama-3 for fast LLMWhisper.cpp for on-device STTGoogle TTS fallback

Build Timeline

Week 1: Foundation and auth

38h
  • Project setup
  • Supabase auth with phone login
  • Basic navigation UI
  • Onboarding flow

Week 2: Voice interface

52h
  • Speech-to-text + TTS integration
  • Voice calibration screen
  • Offline caching system

Week 3: Core AI tutor

48h
  • FastAPI backend
  • LLM prompt engineering
  • Session management

Week 4: Adaptivity engine

45h
  • Mastery tracking logic
  • Adaptive path algorithm
  • Progress database

Week 5: Languages and offline

50h
  • Multi-language support
  • Offline bundle system
  • Testing with local accents

Week 6: Polish and payments

42h
  • Flutterwave integration
  • Guardian reports
  • UI/UX refinement

Week 7: Beta testing prep

35h
  • Internal testing
  • Documentation
  • Analytics implementation

Week 8: Launch readiness

30h
  • App store assets
  • Beta with 100 students
  • Marketing site
Total Timeline: 8 weeks • 435 hours

Pricing Tiers

Free

$0/mo

Limited daily usage, no advanced personalization

  • 15 minutes daily AI tutor
  • Basic offline content
  • Core subjects only

Pro

$5/mo

None

  • Unlimited voice sessions
  • Full adaptive paths
  • All languages
  • Progress tracker

Family

$25/mo

None

  • Everything in Pro
  • Up to 4 student profiles
  • Guardian dashboard
  • Priority support
  • Curriculum alignment

Revenue Projections

MonthUsersConversionMRRARR
Month 11,2004%$600$7,200
Month 614,5006%$13,050$156,600

Unit Economics

$9
CAC
$95
LTV
9%
Churn
78%
Margin
LTV:CAC Ratio: 10.6xExcellent!

Landing Page Copy

Finally, a tutor who never judges and always has time

Learn by speaking. Master subjects at your own pace with a personal AI teacher that understands your accent and your curriculum.

Feature Highlights

Speak naturally in Swahili, English, French & more
Completely private — no classmates watching
Works without internet
Adapts instantly to how you learn
Built for students in Africa and Southeast Asia

Social Proof (Placeholders)

"'I asked the same question 5 times and it never got annoyed. My math is improving.' — Fatou, 13, Senegal"
"'My daughter uses it every evening on our one shared phone. She finally understands science.' — Mama Aisha, Kenya"
"'The voice sounds like my big sister explaining things. Best app ever.' — Budi, 11, Indonesia"

First Three Customers

Partner with two education NGOs (one in Kenya, one in Uganda) to run 6-week pilots in 8 under-resourced schools offering free Family tier access. Use existing teacher WhatsApp groups and education influencers on TikTok to recruit the first 300 beta users. Offer lifetime 50% discount to first 50 parents who upgrade after pilot.

Launch Channels

Product Huntr/EdTechTwitter #EdTechAfricaAfrican teacher Facebook groupsWhatsApp education communitiesLocal radio in Kenya and Indonesia

SEO Keywords

ai voice tutor africaoffline personalized learning appspeak to learn swahili englishjudgment free k12 tutoradaptive ai tutor southeast asia

Competitive Analysis

Khan Academy

khanacademy.org
Free
Strength

Excellent structured video content

Weakness

Requires strong reading skills and is not conversational

Our Advantage

Voice-first + true real-time personalization

Duolingo

duolingo.com
Freemium
Strength

Strong habit formation through gamification

Weakness

Limited to language learning, not core subjects

Our Advantage

Covers math, science, and local curriculum using voice

Elsa Speak

elsaspeak.com
Subscription
Strength

Strong pronunciation feedback

Weakness

Narrow focus on English speaking only

Our Advantage

General academic subjects across multiple languages

🏰 Moat Strategy

Dataset of voice interactions from emerging market students used to fine-tune models for local accents and cultural context. Long-term partnerships with education ministries for verified curriculum alignment.

⏰ Why Now?

Smartphone penetration in Sub-Saharan Africa passed 60% in 2024 while small open-source LLMs and on-device inference became viable, making personalized voice tutoring finally affordable at scale.

Risks & Mitigation

markethigh severity

Extremely low disposable income in target markets

Mitigation

Heavy reliance on freemium + NGO/school sponsorship model for premium features

technicalhigh severity

Poor speech recognition accuracy with strong local accents

Mitigation

Hybrid on-device + cloud approach with text fallback and continuous model improvement

legalmedium severity

Child data privacy regulations across 8+ countries

Mitigation

Minimal data collection, local server options where required, clear consent flows

Validation Roadmap

pre-build12 days

Interview 40 students and 15 teachers in Kenya and Indonesia

Success: 80% confirm voice interface would help them ask questions they avoid in class

mvp28 days

Closed beta with 250 students across 6 schools

Success: 65% daily active users completing 8+ minute sessions for 3+ weeks

launch45 days

Public launch with NGO partners

Success: 3000 downloads and 8% paid conversion in first 45 days

growth60 days

Referral program + TikTok campaign

Success: 15% month-over-month user growth with churn under 10%

Pivot Options

  • B2B school licensing platform with teacher analytics
  • English-only version targeted at global emerging markets
  • Teacher content creation tool using same voice tech

Quick Stats

Build Time
435h
Target MRR (6 mo)
$13,500
Market Size
$9200.0M
Features
9
Database Tables
4
API Endpoints
5