Roadmaps

Interactive learning paths with real prerequisite dependencies. Start with Computer Vision — the complete journey from linear algebra to foundation models.

Available now

Math for ML

The mathematics every ML/Research Engineer is expected to know — linear algebra, calculus, probability & statistics, optimization, and information theory.

6 topics 44h

Python & Scientific Computing

Strong Python plus the scientific stack the labs run on — NumPy, pandas, Matplotlib — and the software-engineering habits that make ML code reliable.

4 topics 20h

Coding & Algorithms

Both coding interview rounds the labs run: classic data structures & algorithms, and ML-flavored implementation — writing layers, losses, and training loops from scratch.

8 topics 47h 30m

Core Machine Learning

The classical ML canon interviews still test — the supervised workflow, tree/kernel models, unsupervised learning, ensembles, and rigorous evaluation.

6 topics 35h 40m

Deep Learning

The general deep-learning core: MLPs, backpropagation, PyTorch, the regularization/normalization toolkit, and the CNN, RNN, and Transformer architectures.

9 topics 53h 40m

ML Systems & Scaling

The engineering that trains and serves frontier models — GPUs and CUDA, distributed training, parallelism (FSDP/ZeRO), mixed precision, and inference optimization.

10 topics 47h 40m

How LLMs Are Trained

From raw text to a helpful assistant — the complete, beginner-first path behind models like Claude, Gemini, and GPT: tokens, transformers, pretraining, and RLHF.

16 topics 65h

Computer Vision

From linear algebra to foundation models — the complete path to mastering modern computer vision.

25 topics 191h

Research Engineering

The craft of the job: reading and reproducing papers, designing and analyzing experiments, debugging models, and communicating results — the core of a Research Engineer role.

9 topics 32h 30m
Planned tracks
Multimodal AI
Vision-language models, CLIP, and cross-modal understanding.
soon
Generative AI
Diffusion, GANs, VAEs, and controllable image synthesis.
soon
Robotics
Perception, control, ROS, and vision for embodied systems.
soon
Autonomous Driving
Sensor fusion, BEV perception, planning, and self-driving stacks.
soon
Medical Imaging
CT/MRI, segmentation, and clinically robust vision models.
soon