Math · code · projects

Master math. Write the code.

Structured courses, interactive lessons, and Python projects from precalc to Calc III, including real quant builds like portfolio optimization.

Start free preview See Pro pricing

The path: launch and beyond

Precalculus

Calculus I

Calculus II

Calculus III

Linear Algebrasoon

Diff Eqsoon

Probabilitysoon

Stochasticsoon

Why MathFlow

Built for people who build things with math.

//Structured paths

Learn in order

Precalc through advanced math, with each course unlocking the next. No random playlist, no guessing what to study.

//Practice that sticks

Checkpoints & drills

Interactive graphs, quizzes, and problem banks built for understanding, not memorizing formulas for a test.

//Math meets code

Python in the browser

Turn functions into simulations and projects. Same concepts you learn in lessons, expressed as real code.

Inside a lesson

See a Precalc lesson in action.

Explanation, moving graph, checkpoint. The same loop every lesson.

Precalc · FoundationsLESSON 01

LIVE

Functions Overview

A function assigns each input exactly one output.

01 Rule: f(x) = x²

02 Shift: g(x) = (x − 0.0)²

03 Checkpoint: evaluate f(3)

f(3) = 9

f(3) = 6

f(3) = 3² + 1

Graph · liveh = 0.0

Transformation preview. Same interactive graphs you get inside the lesson.

Next: domain · range · Python exercise

Trial unlocks full lesson

Python projects

This is why Pro is worth $25/mo.

Free YouTube teaches formulas. MathFlow makes you build them — from a numerical optimization engine to a neural net from scratch in NumPy.

Project · Calc I → NumericsADVANCEDNUMERICAL OPTIMIZATION

x → 4.20

Objective f(x) · GD pathf(4.20) = 58.10

Finite differences estimate f′, then gradient descent walks the path. No optimization libraries. Real numerical methods.

Project brief

Build a numerical optimization engine.

Approximate derivatives, locate and classify critical points, then run gradient descent. Same stack behind modern optimizers, built by hand in Python.

python · guided build

def centered(x, h=1e-4):
    return (f(x+h) - f(x-h)) / (2*h)
 
def gd_step(x, alpha, grad_fn):
    return x - alpha * grad_fn(x)
 
# finite differences → critical points → descent

finite differencescritical pointsgradient descentno SciPy

Open project

Project · Calc III → Deep learningEXPERTNEURAL NET FROM SCRATCH

J → 0.69

Unit circle · decision boundaryepoch 1/7

Points inside vs outside the unit circle. The network learns a nonlinear boundary with tanh → sigmoid — pure NumPy backprop.

Project brief

Build a neural network from scratch.

Forward pass, binary cross-entropy, and full backpropagation by hand. The Calc III chain rule, turned into a working classifier.

python · guided build

Z1 = X @ W1 + b1
A1 = np.tanh(Z1)
Z2 = A1 @ W2 + b2
A2 = sigmoid(Z2)
 
dZ2 = A2 - Y
W2 -= alpha * (A1.T @ dZ2)

forward propbackpropchain ruleno PyTorch

Open project

Featured · Advanced

Build a Numerical Optimization Engine

Finite differences, critical points, and gradient descent. No optimization libraries.

Featured · Expert

Build a Neural Network From Scratch

Forward prop, backprop, and gradient descent in NumPy — no PyTorch. Learn the Calc III math behind deep learning.

Coming soon · Precalc

Compound interest engine

Model loans and growth with functions + Python.

Coming soon · Calc I

Optimization lab

Minimize cost / maximize profit with derivatives.

How it feels

Problem → reasoning → Python.

lesson · practice · codelive

[ Problem ]

▍

[ Reasoning ]

u = x²,  v = sin(x)
u' = 2x, v' = cos(x)
f' = u'v + uv'

[ Python ]

def f_prime(x):
    return 2*x*sin(x) + x**2*cos(x)

# ✓ checkpoint passed

Why pay monthly

Not another course dump. A training system.

The full arc, not a playlist

Precalc → Calc III as one system. Future courses ship into the same Pro plan. You don't rebuy.

Projects that prove the math

Guided Python builds with live labs and checkpoints. Turn lecture ideas into code you can actually run.

Built for careers that pay

Quant, ML, engineering. Monthly access means you keep drilling until interviews and coursework stop being scary.

$25 vs. getting stuck alone

Less than a tutoring hour. Lessons, exams, and Python projects included. Cancel anytime if it's not for you.

Pro · $25/mo or $20/mo annual

Everything unlocks. New courses stay included.

Start free trial

Preview only · full access in Pro

Ready to start learning? Try free first.

First 4 lessons per course and practice are free. Upgrade to Pro for the full path, exams, and guided Python projects — with a 3-day trial.

Start free preview →full pricing breakdown

Built by @jaimecodes