Curriculum & parent guide
From the math up: really understand AI
This is the complete map of what a Math to Machine student learns — 78 weeks, six phases, from school-level algebra to shipping an agentic-AI capstone. It's honest and self-contained; keep it, print it, or share it with a teacher. The first five lessons are free, no card needed.
- 78 weeks
- 6 phases, one path
- 234 lessons
- every one hands-on
- 3 tracks
- Class 9-10 / 11-12 / Grad
- ₹399/mo
- or ₹2,000/year
How every lesson works
Not a video archive. Every concept follows the same evidence-backed loop: you predict before the rule is shown, learn it, build something small with your own hands, then explain it back in your own words before a quick check you can't skip. Each week ends with a boss check at 80% to mark it mastered. An AI tutor gives hints, never the full answer.
This isn't a hunch: active learning (Freeman et al., 2014), the testing effect (Roediger & Karpicke, 2006), and the generation effect (Slamecka & Graf, 1978) all point the same way — you learn by doing and retrieving, not watching.
The 78-week curriculum
Each phase builds on the last; nothing is skipped.
Phase 1: Maths & Programming Foundations
Weeks 1–20The bedrock. School-level maths made rigorous, plus the Python and tooling every later phase assumes.
- Set theory & logic (the language of proof)
- Vectors, matrices, eigenvalues & SVD
- Derivatives, gradients & the chain rule
- Multivariable calculus & optimisation (incl. Lagrange multipliers)
- Probability, common distributions & Bayes' theorem
- Statistics: sampling, the Central Limit Theorem, MLE, confidence intervals
- Python, NumPy, and Git
Phase 2: Classical Machine Learning
Weeks 21–32The workhorse algorithms — and, more importantly, how to evaluate them honestly on real data.
- Linear & logistic regression
- Decision trees, random forests & ensembles
- Support vector machines & the kernel trick
- k-means & clustering
- PCA & dimensionality reduction
- Model evaluation, cross-validation & the bias–variance trade-off
Phase 3: Deep Learning
Weeks 33–48Neural networks from a single neuron up to the transformer that powers modern AI.
- What a single neuron computes; activation functions
- Backpropagation, derived and computed by hand
- Convolutional networks (CNNs) for images
- Recurrent networks (RNNs) & sequences
- Transformers & the attention mechanism
- Regularisation and the real-world training loop
Phase 4: Generative AI
Weeks 49–62How today's language and image models actually work — no hand-waving.
- Tokenization & embeddings
- Large language models & next-token prediction
- Diffusion models (how AI paints from noise)
- Retrieval-augmented generation (RAG)
- Multimodal models
- Fine-tuning and adaptation
Phase 5: Agentic AI
Weeks 63–72Turning a model into a system that can plan, use tools, and act.
- Tool use & function calling
- Planning & reasoning loops
- Memory & long context
- The Model Context Protocol (MCP) & connectors
- Multi-agent orchestration
- Evaluation & guardrails
Phase 6: Production & Capstone
Weeks 73–78Ship it. Take one idea from notebook to something that runs and is monitored.
- MLOps & experiment tracking
- Serving models behind an API
- Monitoring & data drift
- Cost & latency in the real world
- A faculty-style reviewed capstone project
Three depth-tuned tracks
Same curriculum, depth adjusted to where a student actually is. You pick the track; you can change later.
Class 9–10
Intuition-first. Friendlier framings, more scaffolding, the same concepts introduced at an accessible depth.
Class 11–12
Conceptual depth with the full maths, tuned to a senior-secondary student's footing.
Graduation / BE1
Code-first and rigorous, with heavier build tasks and the complete derivations.
For parents
Built to keep your child safe and on track.
Parent-verified for under-16
No account is created for a child under 16 until you approve it via a one-time code sent to you. You can revoke access anytime.
Weekly progress emails
A Monday summary of what your child learned, where they got stuck, and how much time they spent. No surprises.
No DMs, no social, no ads
Under-16 accounts cannot message other learners. There are no advertisements, and we never sell data.
DPDP-compliant, India-first
We follow the Digital Personal Data Protection Act, 2023. Every policy is in plain English on our /legal page.
Maps to CBSE / ICSE chapters
Lessons surface the school-board chapter they cover, so studying for AI also reinforces school maths.
7-day no-questions refund
Cancel within 7 days of subscribing for a full refund. We don't auto-charge during this launch.
Pricing
Try the first five lessons free — no card. Subscribe only when you're sure.
| Plan | Price | What you get |
|---|---|---|
| Free | ₹0 | First 5 lessons, 5 AI-tutor messages/day |
| Premium Monthly | ₹399 / month | All 234 lessons, 100 tutor messages/day, cohorts, portfolio publishing, completion certificate |
| Premium Yearly | ₹2,000 / year | Everything in Monthly — ≈58% cheaper (about ₹167/month) |
All prices in INR, GST inclusive. 7-day refund for new subscribers.
Ready to try it?
Reading about it isn't the same as doing it. The first five lessons are free — no card, no commitment.
Questions? Visit mathtomachine.com or write to support@mathtomachine.com.