AI Learning Roadmaps

Discover the right path for you to learn AI, and explore the techniques, algorithms, frameworks and tools through interactive knowledge maps that show you what to learn, in what order, and the skills you need to move to the next level.

41Roadmaps
820Topics
10Specializations
82Techniques and tools

Map categories

Explore your path

Choose the goal closest to you, and the matching map will open with a suggested first step.

Code productively and safely with AI AI-Assisted Coding (Vibe Coding) — for programmers, researchers and computing students who know the basics of programming Become an AI engineer AI Engineer — for beginners and developers Master machine learning Machine Learning — for those who know Python basics Specialise in deep learning Deep Learning — for those who have mastered machine learning basics Build with LLMs Generative AI & LLM — for those who know Python and deep learning basics Become an AI researcher AI Researcher — for graduate students and researchers Build reliable AI agents AI Agents — for those fluent in Python who know language model basics Specialise in computer vision and medical imaging Computer Vision & Medical Image Analysis — for those who know deep learning basics and PyTorch Understand the maths behind AI Mathematics & Statistics for AI — for those who have studied secondary-school or first-year university mathematics Master the AI programming toolkit Programming & AI Frameworks — for programming beginners or those coming from other languages Specialise in NLP, including Arabic Natural Language Processing (NLP) — for those who know Python and machine learning basics Explain and trust AI models Explainable AI (XAI) — for those who train machine learning or deep learning models Run ML reliably in production MLOps — for those who train models and know the basics of Git and Docker Enter quantum ML on solid ground Quantum Machine Learning (Quantum ML) — for those who have mastered linear algebra and machine learning basics Complete and publish a graduate AI thesis AI for Graduate Students — for master's and PhD students and their supervisors Work at the intersection of AI and security AI & Cybersecurity — for those who know machine learning and deep learning basics Master data analysis and data pipelines Data Science & Data Engineering — for data beginners and those preparing to learn machine learning Deploy AI models on edge devices Edge AI & TinyML — for those who know deep learning and PyTorch or TensorFlow Understand RL from MDPs to RLHF and GRPO Reinforcement Learning — for those who know probability and neural network basics Build reliable forecasts Time Series & Forecasting — for those who know basic statistics and pandas Learn GNNs and their applications Graph Neural Networks (GNNs) — for those who know linear algebra, neural networks and PyTorch Build fair, responsible and compliant AI AI Ethics & Governance — for developers, researchers and decision-makers in organizations Become a machine learning engineer Machine Learning Engineer — for developers and computing graduates Become a data scientist Data Scientist — for beginners in data and graduates in statistics, computing and engineering Become a medical AI researcher Medical AI Researcher — for postgraduate students, physicians and biomedical engineers Master Claude Code Learning Claude Code roadmap — for programmers and researchers who know Git and the command line Master OpenAI Codex Learning Codex roadmap — for programmers and researchers who know Git and the command line Work productively with Antigravity Learning Antigravity roadmap — for programmers who know Git and want an agentic development platform Master Python for AI Python for AI roadmap — for programming beginners and for researchers who write Python without engineering foundations Learn cybersecurity from the ground up Cybersecurity learning map — for students, programmers and system administrators new to security Become an ethical hacker Ethical hacking learning roadmap — for those who have completed the basics of Linux and networking and want to specialize in penetration testing Fine-tune, evaluate and deploy LLMs LLM Engineering — for those who know Python, PyTorch and the basics of transformers Build accurate RAG systems Retrieval-Augmented Generation (RAG) — for those who know Python and the basics of language models and embeddings Understand and adapt multimodal models Multimodal Models (Multimodal / VLM) — for those who know deep learning, transformers and the basics of computer vision Build recommender systems Recommender Systems — for those who know the basics of machine learning and Python Train models without centralising data Federated Learning — for those who know deep learning, PyTorch and the basics of privacy Build AIoT systems AI for the Internet of Things (AIoT) — for those who know Python and the basics of machine learning and want to work with hardware Master Siamese networks Siamese Networks — for those who know convolutional neural networks and PyTorch Go deep into quantum AI research Advanced Quantum AI — for those who have completed the «Quantum Machine Learning» map and know linear algebra and machine learning well Understand capsule networks Capsule Networks — for those who know convolutional networks and PyTorch Build reliable data pipelines for AI AI Data Engineering — for those who know Python, SQL and the basics of machine learning

All maps

AI-Assisted Coding (Vibe Coding) AI-Assisted Coding (Vibe Coding) How to work productively and safely with AI coding tools such as Claude Code, Cursor, Antigravity and Copilot: foundations, agentic workflows, context files… Programming and framework maps Intermediate 31 topics 6–10 weeks at 6 hours a week AI Engineer AI Engineer A complete career path that takes you from computing, programming and mathematics fundamentals to machine learning, deep learning, language models, intelligent agents and deployment, ending with five… Career specialization maps Intermediate 81 topics 12–18 months at 10 hours a week Machine Learning Machine Learning A classical machine learning roadmap from the mathematical foundations and data preparation to supervised, unsupervised and ensemble algorithms, with correct evaluation and hyperparameter tuning… Machine learning maps Intermediate 49 topics 5–7 months at 8 hours a week Deep Learning Deep Learning From the perceptron and backpropagation to convolutional and recurrent networks, transformers, generative models, graph networks and self-supervised learning, with frameworks… Deep learning maps Advanced 36 topics 6–9 months at 10 hours a week Generative AI & LLM Generative AI & LLM How large language models work, from attention and the transformer to pre-training, instruction tuning and LoRA, then prompt engineering, function calling, RAG and evaluation… Generative AI maps Advanced 33 topics 4–6 months at 8 hours a week AI Researcher AI Researcher The AI research scientist path: mathematical foundations, literature review and identifying the gap, designing fair experiments and baselines, methodological innovation… Scientific research maps Advanced 29 topics 9–12 months alongside a research project AI Agents AI Agents From language model foundations to tool calling, agentic design patterns, retrieval, frameworks and multi-agent systems, with security, evaluation and monitoring. AI agent maps Advanced 39 topics 3–5 months at 10 hours a week Computer Vision & Medical Image Analysis Computer Vision & Medical Image Analysis From digital images and classical processing to convolutional networks, transfer learning, detection, segmentation and vision foundation models, then medical imaging: imaging modalities… Applied AI maps Advanced 28 topics 5–7 months at 10 hours a week Mathematics & Statistics for AI Mathematics & Statistics for AI Linear algebra and matrix decompositions, calculus and automatic differentiation, probability and statistical inference, optimization, information theory and numerical computing as… Mathematics and statistics maps Intermediate 31 topics 4–6 months at 8 hours a week Programming & AI Frameworks Programming & AI Frameworks Python for AI, developer tools (Git, the command line, Jupyter and package management), scientific libraries, deep learning frameworks, language model libraries… Programming and framework maps Beginner 32 topics 3–5 months at 8 hours a week Natural Language Processing Natural Language Processing Text preprocessing and the specifics of Arabic, text representation, sequence models and attention, pre-trained models, and core language tasks from classification to… Applied AI maps Intermediate 33 topics 4–6 months at 10 hours a week Explainable AI (XAI) Explainable AI (XAI) Foundations of interpretability and inherently interpretable models, model-agnostic methods (SHAP, LIME and counterfactual explanations), interpreting deep networks (Grad-CAM… Advanced technology maps Advanced 22 topics 3–4 months at 8 hours a week MLOps MLOps MLOps principles, data versioning and validation, experiment tracking and the model registry, pipelines, CI/CD and machine learning tests, serving and optimization… Advanced technology maps Advanced 28 topics 3–5 months at 10 hours a week Quantum Machine Learning Quantum Machine Learning Foundations of quantum computing, variational algorithms and parameter-shift gradients, data encoding, variational classifiers and quantum kernels, NISQ hardware noise and plateaus… Advanced technology maps Expert 21 topics 4–6 months at 8 hours a week AI for Graduate Students AI for Graduate Students A practical path for master's and PhD students: choosing the research problem, writing the proposal, ethics approvals, designing repeatable experiments, and publishing in journals… Scientific research maps Intermediate 29 topics Throughout the first research year, alongside the thesis AI & Cybersecurity AI & Cybersecurity Both directions together: machine learning for cyber defence (intrusion and anomaly detection), and the security of AI systems themselves (adversarial examples, data poisoning, attacks… Advanced technology maps Advanced 24 topics 3–5 months at 8 hours a week Data Science & Data Engineering Data Science & Data Engineering From cleaning, analysing and visualizing data to SQL and modern tools (Polars and DuckDB), then warehouses, dbt transformations and scheduled data pipelines with Airflow… Mathematics and statistics maps Beginner 29 topics 4–6 months at 8 hours a week Edge AI & TinyML Edge AI & TinyML Running models on the devices themselves: measuring latency, memory and energy; quantization, pruning, distillation and efficient architectures; then deployment… Advanced technology maps Advanced 25 topics 3–4 months at 8 hours a week Reinforcement Learning Reinforcement Learning From Markov decision processes (MDP) and Q-Learning to DQN, policy gradients and PPO, all the way to RLHF, DPO and GRPO in training language models, as in DeepSeek-R1. Machine learning maps Advanced 18 topics 3–4 months at 8 hours a week Time Series & Forecasting Time Series & Forecasting From decomposition, stationarity, exponential smoothing and ARIMA, through backtesting and gradient boosting with lag features, to time-series Transformer models (Informer and TFT… Applied AI maps Intermediate 18 topics 3–4 months at 6 hours a week Graph Neural Networks (GNNs) Graph Neural Networks (GNNs) From graph basics and node embeddings to message passing, GCN, GraphSAGE, GAT, GIN and the PyTorch Geometric library, then applications in molecules and discovery… Deep learning maps Advanced 19 topics 2–3 months at 8 hours a week AI Ethics & Governance AI Ethics & Governance International principles (UNESCO and OECD), sources of bias and documented examples, fairness metrics, the risks of language models, and governance through the act… Advanced technology maps Intermediate 18 topics 6–10 weeks at 5 hours a week Machine Learning Engineer Machine Learning Engineer A path by job role: production-quality software, the core of modelling and evaluation, experiment tracking and training pipelines, serving and containers, continuous integration and monitoring; that is… Career specialization maps Intermediate 37 topics 9–12 months at 10 hours a week Data Scientist Data Scientist A path by job role: Python and SQL, statistics and experimental design and A/B tests, cleaning, exploratory analysis and data storytelling, and predictive modelling… Career specialization maps Intermediate 29 topics 8–12 months at 8 hours a week Medical AI Researcher Medical AI Researcher A path by job role: medical data, its formats and the ethics of using it, deep learning for medical images, external validation and calibration, the established reporting guidelines… Career specialization maps Advanced 29 topics About 12 months alongside a medical research project Learning Claude Code Learning Claude Code Learn Anthropic's Claude Code based on the official documentation only: installation and sessions, CLAUDE.md memory, settings and permissions, sandboxing and protection against injection… Programming and framework maps Intermediate 35 topics 4–6 weeks at 5 hours a week Learning OpenAI Codex Learning OpenAI Codex Learn OpenAI's Codex coding agent based on the official documentation only: the CLI and editor extension, AGENTS.md files, the config.toml settings file… Programming and framework maps Intermediate 30 topics 3–5 weeks at 5 hours a week Learning Google Antigravity Learning Google Antigravity Learn the Google Antigravity platform: choosing between the app, the agy CLI, the editor, the SDK and the extensions, installation, workflows with agents, integrations, and the log… Programming and framework maps Intermediate 26 topics 3–4 weeks at 5 hours a week Python for AI Python for AI Python for AI work: the language in depth, environments and quality tools (PEP 8, Ruff, type hints, Pydantic and pytest), numerical Python and operations… Programming and framework maps Beginner 29 topics 3–4 months at 6 hours a week Cybersecurity Learning Path Cybersecurity Learning Path A general path in cybersecurity: principles and threat modelling, Linux, networking and traffic analysis, applied cryptography and identity, OWASP Top 10:2025 and secure coding… Advanced technology maps Beginner 23 topics 6–9 months at 8 hours a week Ethical Hacking Learning Path Ethical Hacking Learning Path Ethical hacking and penetration testing: permission, scope and responsible disclosure, methodology (NIST SP 800-115), reconnaissance and enumeration, web application testing with the guide… Advanced technology maps Intermediate 32 topics 6–12 months at 8 hours a week LLM Engineering: Fine-tuning, Evaluation and Deployment LLM Engineering: Fine-tuning, Evaluation and Deployment Large language model engineering from start to deployment: choosing the base model, preparing fine-tuning data, tuning with LoRA and QLoRA, preference alignment… Generative AI maps Advanced 34 topics 4–6 months at 8 hours a week Retrieval-Augmented Generation (RAG) Retrieval-Augmented Generation (RAG) Building retrieval-augmented generation systems for production: extracting text from documents and chunking it, embeddings, vector databases and their indexes (HNSW, pgvector and Qdrant)… Generative AI maps Advanced 32 topics 3–4 months at 8 hours a week Multimodal Models and Vision–Language Models Multimodal Models and Vision–Language Models Multimodal AI: modality encoders and ways of fusing them, contrastive vision-language training (CLIP, ALIGN and SigLIP), vision-language model architectures… Generative AI maps Advanced 27 topics 3–5 months at 8 hours a week Recommender Systems Recommender Systems Recommender systems from foundations to production: explicit and implicit feedback, content-based and collaborative filtering, matrix factorization, BPR and factorization machines… Applied AI maps Intermediate 28 topics 3–4 months at 8 hours a week Federated Learning Federated Learning Federated learning: training shared models without collecting the data in one place. Cross-device and cross-organization learning, the FedAvg algorithm and heterogeneous data, and… Advanced technology maps Advanced 29 topics 3–4 months at 8 hours a week AI for the Internet of Things (AIoT) AI for the Internet of Things (AIoT) AI in the Internet of Things from sensor to cloud: system layers, sensors and microcontrollers, the MQTT and CoAP protocols, and series data pipelines… Applied AI maps Intermediate 28 topics 3–5 months at 8 hours a week Siamese Networks and Similarity Learning Siamese Networks and Similarity Learning Siamese networks and deep metric learning: the difference between learning similarity and classification, twin encoders with shared weights, building pairs and triplets, and the loss… Deep learning maps Intermediate 30 topics 2–3 months at 6 hours a week Advanced Quantum AI Advanced Quantum AI An advanced map that complements the «Quantum Machine Learning» map: the algorithms behind speed-up claims (phase estimation, Grover and HHL) and dequantization, and data re-uploading… Advanced technology maps Expert 34 topics 4–6 months at 8 hours a week Capsule Networks Capsule Networks Capsule networks: why pooling loses pose information, the difference between invariance and equivariance, capsule vectors and the squash function, part-to-whole transformation matrices… Deep learning maps Advanced 30 topics 2–3 months at 6 hours a week AI Data Engineering AI Data Engineering Data engineering for AI systems: the data life cycle, the Parquet and Arrow formats and data lake tables, and reproducible batch and streaming pipelines… Advanced technology maps Intermediate 36 topics 3–5 months at 8 hours a week