TutorAI
- Github: https://github.com/dan-the-man639/tutorAI
- Linkedin: https://www.linkedin.com/in/danny-yang639/
- Video Demo: https://youtu.be/NehwRFE2Res?si=GcNgIAa7dGwLDg2R
Problem Statement
- Core issue: Online learning platforms still push identical videos to every learner, while private tutoring is costly and scarce.
- What’s broken / not working
- Static videos cannot adapt to a learner’s pace or address individual gaps.
- Massive-open courses average completion rates below 10 %.
- Private STEM tutors cost US $60–150 per hour, pricing many families out.
- Static videos cannot adapt to a learner’s pace or address individual gaps.
- Why this persists
- Existing platforms were built for broadcast rather than conversation.
- Real-time visual explanations once required skilled humans; only recent AI + GPU advances make automation viable.
- Existing platforms were built for broadcast rather than conversation.
Solution -> TutorAI
- What it is: A voice-first AI tutor that listens, explains with live animations, and obeys playback commands spoken by the learner.
- How we solve the problem
- GPT-4o reasons through any math or STEM prompt.
- A cloud engine renders bespoke 10-second animations (e.g., Riemann sums collapsing into an integral).
- ElevenLabs voices the explanation; a lip-synced avatar boosts engagement.
- A JSON command layer obeys natural speech like “rewind,” “slow to half-speed,” or “highlight step three.”
- GPT-4o reasons through any math or STEM prompt.
- Use-case snapshots
- AP-Calculus student: “Show how Riemann sums become an integral.”
- Linear-algebra freshman: “Visualise eigenvectors of this 2×2 matrix.”
- Data-analyst up-skiller uploads a PDF and receives a spoken walkthrough of each formula.
- AP-Calculus student: “Show how Riemann sums become an integral.”
Market & Target Demographic Research
- Total Addressable Market (TAM): Global e-learning plus private tutoring ≈ US $485 B (2025).
- Serviceable Available Market (SAM): English-language, online STEM learners ≈ US $95 B.
- Serviceable Obtainable Market (SOM): Capturing 0.12 % over three years ≈ US $115 M in annual recurring revenue.
- Ideal early customers: North-American high-school and college STEM learners, self-taught programmers, and tutoring centers seeking scalable help.
Business Model
- How we make money
- Freemium with a US $15 / month TutorAI Pro tier.
- B2B SaaS: US $5 per seat per month for schools and tutoring centers.
- API licensing for publishers embedding TutorAI explanations.
- Freemium with a US $15 / month TutorAI Pro tier.
- Revenue mechanics
- Subscriptions create predictable ARR.
- Volume licenses yield low-churn institutional revenue.
- White-label API earns per-lesson transaction fees.
- Subscriptions create predictable ARR.
- Extra details
- Gross margin exceeds 80 % after GPU and token costs (~ US $0.06 per full lesson).
- Planned partnerships with cloud-GPU vendors and curriculum publishers.
- Gross margin exceeds 80 % after GPU and token costs (~ US $0.06 per full lesson).
Traction
- Alpha prototype live: typed prompt → custom animation; voice layer specification complete.
- Fifty closed-beta learners averaging 12 minutes daily use.
- Two tutoring-center chains have signed letters of intent for pilot seats.
- Roadmap
- Q3 2025: launch web MVP with full voice and video-control commands.
- Q4 2025: reach 1 000 paying users and five pilot schools; raise US $1.5 M seed round.
- 2026: release AR-glasses demo and add multilingual support.
- Q3 2025: launch web MVP with full voice and video-control commands.
Risk Mitigation & MVP
- Technical cost risk – mitigated with caching and shorter default animations.
- Pedagogical accuracy – human educator review loop plus user feedback buttons.
- LLM hallucinations – guard-rail prompts and post-processing validators.
- Adoption friction – demo requires no login; voice UI lowers the barrier.
- MVP definition – web app featuring mic input, GPT-generated text, live animations, ElevenLabs voice, and six voice commands (play, pause, rewind, slow_down, speed_up, highlight_step) delivered in under 25 seconds end-to-end.
One-Pager Summary
TutorAI is the affordable AI-powered STEM tutor that fuses GPT-4o reasoning, real-time animations, and natural voice interaction. Learners simply speak; TutorAI explains, visualises, and obeys commands like “rewind to step three.” Targeting a slice of the US $485 B e-learning and tutoring market, TutorAI starts with a freemium model, institutional licenses, and API sales, aiming for US $115 M ARR within three years. Alpha prototype is live with 50 active testers and tutoring-center LOIs. Next milestone: Q3 2025 MVP with complete voice control. TutorAI delivers 3Blue1Brown-level clarity with Jarvis-style interactivity—at less than 10 % of human-tutor cost.
Log in or sign up for Devpost to join the conversation.