AI · Robotics · Control
Learn the hard parts
properly.
Self-paced courses on deep learning, robot dynamics and control theory — derived from first principles, not skimmed from a framework's docs.
Three tracks
Pick a track, or work through all three.
Artificial Intelligence
Backpropagation by hand, then the architectures that matter. Reinforcement learning from Markov decision processes to policies that survive contact with hardware.
Robotics
Kinematics and dynamics derived rather than imported. State estimation and sensor fusion that does not quietly lie to you about where the robot is.
Control
Optimal control from the Riccati equation outward, and model predictive control with real constraints, real time budgets and real warm starts.
Courses
Six courses. No filler.
Each course is self-contained. Taken together they form a path from the mathematics to a controller running on real hardware.
How we teach
Derive it once.
Then you own it.
Every course starts from the maths and builds up. You will implement the algorithm before you are shown a library that wraps it — because the library is easy and the maths is not.