# AI-ML-Roadmap-from-scratch: A Free 11-Module Learning Path > AI-ML-Roadmap-from-scratch is a curated GitHub repository that organises free learning resources for artificial intelligence, machine learning, deep learning, generative AI, and related fields into a structured 11-module progression, targeting learners who want a zero-to-skilled path without paid courses. **aadi1011/AI-ML-Roadmap-from-scratch** — Become skilled in Artificial Intelligence, Machine Learning, Generative AI, Deep Learning, Data Science, Natural Language Processing, Reinforcement Learning and more with this complete 0 to 100 repository. - Repository: https://github.com/aadi1011/AI-ML-Roadmap-from-scratch - Website: https://aadi1011.github.io/AI-ML-Roadmap-from-scratch/ - Stars: 4,216 · Forks: 784 - Language: Unknown - License: MIT - Published: 2026-09-10 · Updated: 2026-09-10 · Language: en - Canonical page: https://hysenlabs.com/projects/aadi1011-ai-ml-roadmap-from-scratch ## What This Repository Is and Who It Serves AI-ML-Roadmap-from-scratch is a curated index of free online resources, not a textbook or a code project. It organises YouTube playlists, MOOC courses, and documentation links into a ranked progression from absolute beginner to advanced practitioner. The README describes it as a "complete 0 to 100 repository" covering artificial intelligence, machine learning, generative AI, deep learning, data science, natural language processing, and reinforcement learning. The target audience is a self-directed learner who is starting from little or no background. Module 0 assumes the reader needs to install Python and a code editor before anything else. Module 10 covers agentic AI, which presupposes fluency with everything in the preceding modules. The curve is steep, but the README notes that modules can be followed simultaneously as well as in order. The resources are drawn from free tiers of well-known platforms: YouTube, edX, Coursera (free audit), Codecademy, W3Schools, HackerRank, and NPTEL. Items marked with a star are described by the README as highly recommended. The emphasis is on availability at no cost: every link in the main learning pathway is free to access, though some courses offer paid certificates. ## How the 11 Modules Are Organised The repository uses a flat structure. The README is the primary navigation surface. The resources/ directory holds supplementary material, and Packages.md lists common Python libraries used in AI and ML work. There are no notebooks, no code exercises, and no automated tests: this is a link collection with editorial curation. The module sequence follows a deliberate order of dependencies: - Module 0: Tooling setup (Python, VS Code, pip, common packages) - Module 1: Mathematics (linear algebra from MIT OCW, discrete mathematics from NPTEL, statistics) - Module 2: Python programming foundations (MIT and Harvard intro courses, HackerRank practice) - Module 3: Data science with Python - Module 4: Machine learning - Module 5: Computer vision - Module 6: Deep learning and neural networks - Module 7: Generative AI, including Sub-Module 7A on retrieval-augmented generation - Module 8: Natural language processing - Module 9: Reinforcement learning - Module 10: Agentic AI A bonus module covers advanced learning pathway courses beyond the main sequence. The README also lists projects, interesting websites, AI newsletters, and AI blogs as supplementary sections beyond the module path. ## Starting at Module 0 and Moving Through Math Module 0 is where the README tells learners to begin. It links to Python 3.14, VS Code, a pip installation guide on GeeksforGeeks, and the repository's own Packages.md file, which catalogs the common Python libraries used across ML workflows. Module 1 addresses the mathematical prerequisites directly, before any code. The README includes an MIT OpenCourseWare linear algebra lecture series, an NPTEL discrete mathematics course (marked as highly recommended), and a Codecademy path on fundamental math for data science. The reasoning is practical: learners who skip this module often encounter unexplained notation in ML papers and courses later. The README does not require completing all of Module 1 before starting Module 2; it notes modules can run in parallel. Module 2 builds Python skill through the MITx and HarvardX introduction courses on edX, W3Schools for quick reference, a four-hour YouTube course, and HackerRank practice problems (also marked highly recommended). The HackerRank Python basic certification is listed as an optional milestone. The repository's companion YouTube content, linked in the README, provides a video walkthrough of the resources in Part 1 and Part 2 videos by Aadith Sukumar. ## Advanced Modules: Generative AI, RAG, and Agentic AI The later modules address the subjects that have become the most active hiring and research areas. Module 7 on generative AI is expanded with Sub-Module 7A specifically on retrieval-augmented generation, reflecting RAG's prominence as an implementation pattern. The sub-module exists as a nested entry rather than a separate module, which keeps the overall progression clear while giving RAG its own curated resources. Module 10, agentic AI, is the terminus of the main pathway. Adding it signals that the curator treats agent-based systems as a distinct skill set beyond classical ML, rather than a subset of generative AI. Module 9 on reinforcement learning appears after NLP in Module 8, which is a departure from some other curricula that place RL earlier. The ordering reflects the practical reality that most ML jobs involve supervised or generative techniques first, and RL is a specialisation learned after the foundations are solid. The PROJECTS section in the README is listed as a supplementary item after the modules but is not elaborated in the portion of the README available for review. It indicates that practical project work is treated as a companion to the coursework rather than integrated into the module sequence. ## Coverage Gaps and Cases Where This Repository Is the Wrong Choice The repository is a link collection, which means its value depends on the quality and availability of the external sources it points to. The README does not document what happens when a linked course is retired, a YouTube playlist is removed, or a free tier becomes paid. Some links in a roadmap of this size will inevitably go stale. The repository does not contain practice problems, project starters, or code examples of its own. A learner who needs immediate feedback on whether their understanding is correct must find that feedback inside the linked courses. There is no built-in assessment mechanism. For learners who need an employer-recognised credential, the repository links to Coursera and edX courses that offer paid certificates, but the repository itself does not issue any certification. The free audit path through most of these courses does not include grading. The repository also does not cover MLOps, model deployment, monitoring, or the engineering infrastructure around production ML systems. Those topics are absent from the module list. A learner finishing Module 10 will have studied the techniques but not the production practices. ## roadmap.sh as a Structured Alternative roadmap.sh is a well-known open-source project that publishes interactive, community-maintained learning roadmaps for software engineering topics including machine learning and AI. The key difference in approach is presentation: roadmap.sh renders each roadmap as a visual graph of skills and concepts, with nodes that mark a learner's progress. AI-ML-Roadmap-from-scratch uses a sequential module list in a README, which is easier to read linearly but does not show the dependency graph visually. roadmap.sh covers a broader set of engineering topics and is maintained by a larger community with regular pull requests. AI-ML-Roadmap-from-scratch is more opinionated in its resource selection, curated by a single author who has also published companion YouTube videos explaining the choices. For learners who prefer a personal curation with a single point of accountability, the single-author model has an advantage: changes are deliberate rather than the result of community consensus. The two projects are complementary rather than mutually exclusive, but a learner who wants to see how topics relate across the field before committing to a linear sequence may find roadmap.sh more useful as a starting orientation. ## Maintenance and Licence The last push to the repository was on 2026-08-12. The project is MIT licensed, which permits free use, copying, and redistribution. The repository has no GitHub releases and tracks changes through direct commits to main. Because the repository's value is in the links it curates, the relevant maintenance question is not code stability but link freshness. The absence of an automated link-checking workflow (no CI configuration is visible in the top-level entries) means broken links are discovered manually, through issues, or through a contributor noticing. The CONTRIBUTING.md file suggests the repository accepts community contributions. ## Conclusion AI-ML-Roadmap-from-scratch suits self-directed learners who want a pre-ordered sequence of free resources and do not want to spend time finding and vetting individual courses. It is not the right choice for someone who needs assessed certification, hands-on project grading, or a structured cohort. Before committing to it, verify that Module 0's tooling links (Python, VS Code, pip) still resolve, since external URLs in a curated list can go stale and the repository's last push was on 2026-08-12. ## FAQ ### Can you provide a roadmap for learning AI from scratch? The AI-ML-Roadmap-from-scratch repository provides exactly that: an 11-module sequence starting from Python setup and mathematics (Module 0 and 1) and progressing through machine learning, deep learning, generative AI, NLP, reinforcement learning, and agentic AI. The README notes that modules can be followed in parallel as well as in order. ### Can I learn AI ML on my own? The repository is built for exactly that purpose. All resources in the main module sequence are free to access, covering YouTube playlists, NPTEL and MIT OpenCourseWare courses, and practice sites like HackerRank. The README does not require any prior AI or ML background before Module 0. ### Does the repository include hands-on project work or coding exercises? The repository itself does not contain code exercises or project starters. It links to external courses and platforms that include practice, and lists a PROJECTS section as supplementary content. Hands-on work happens inside the linked platforms, not in the repository itself. ## Sources - [aadi1011/AI-ML-Roadmap-from-scratch on GitHub](https://github.com/aadi1011/AI-ML-Roadmap-from-scratch) - [Issues](https://github.com/aadi1011/AI-ML-Roadmap-from-scratch/issues) - [License: MIT](https://github.com/aadi1011/AI-ML-Roadmap-from-scratch/blob/main/LICENSE) - [Project website](https://aadi1011.github.io/AI-ML-Roadmap-from-scratch/) - [README](https://github.com/aadi1011/AI-ML-Roadmap-from-scratch/blob/main/README.md) --- Hysen Labs editorial analysis, written from the project's own repository and release notes. Cite the canonical page: https://hysenlabs.com/projects/aadi1011-ai-ml-roadmap-from-scratch