/ Home

GenAI Resources:

Note: tbw

1

𝟏𝟐 𝐌𝐮𝐬𝐭-𝐅𝐨𝐥𝐥𝐨𝐰 𝐆𝐢𝐭𝐇𝐮𝐛 𝐑𝐞𝐩𝐨𝐬𝐢𝐭𝐨𝐫𝐢𝐞𝐬 𝐟𝐨𝐫 𝐀𝐈 𝐄𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐬 If you’re learning AI or building your first production-grade system, GitHub can be your best teacher.

Thanks, Sivasankar, for curating this list…

𝐇𝐞𝐫𝐞’𝐬 𝐚 𝐜𝐚𝐭𝐞𝐠𝐨𝐫𝐢𝐳𝐞𝐝 𝐫𝐨𝐚𝐝𝐦𝐚𝐩 𝐭𝐨 𝐡𝐞𝐥𝐩 𝐲𝐨𝐮 𝐠𝐫𝐨𝐰 𝐟𝐫𝐨𝐦 𝐁𝐞𝐠𝐢𝐧𝐧𝐞𝐫 → 𝐈𝐧𝐭𝐞𝐫𝐦𝐞𝐝𝐢𝐚𝐭𝐞 → 𝐀𝐝𝐯𝐚𝐧𝐜𝐞𝐝:

𝐁𝐞𝐠𝐢𝐧𝐧𝐞𝐫 𝐋𝐞𝐯𝐞𝐥 𝐁𝐮𝐢𝐥𝐝 𝐘𝐨𝐮𝐫 𝐅𝐨𝐮𝐧𝐝𝐚𝐭𝐢𝐨𝐧

1. AI-For-Beginners (Microsoft): A complete 12-week curriculum covering AI fundamentals from classical AI to ethics. Link: https://github.com/microsoft/AI-For-Beginners

2. ML-For-Beginners (Microsoft): End-to-end introduction to ML with projects, visual explanations, and Jupyter notebooks. Link: https://github.com/microsoft/ML-For-Beginners

3. Generative-AI-For-Beginners (Microsoft): A modern guide to generative AI and LLM fundamentals your entry into real-world GenAI. Link: https://github.com/microsoft/generative-ai-for-beginners

4. AI-Agents-For-Beginners (Microsoft): Learn the principles of AI agents and how they work in real-world scenarios. Link: https://github.com/microsoft/ai-agents-for-beginners

𝐈𝐧𝐭𝐞𝐫𝐦𝐞𝐝𝐢𝐚𝐭𝐞 𝐋𝐞𝐯𝐞𝐥 𝐇𝐚𝐧𝐝𝐬-𝐨𝐧 𝐏𝐫𝐚𝐜𝐭𝐢𝐜𝐞 & 𝐄𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐢𝐧𝐠

1. Learn-AI-Engineering: Curated roadmap covering AI/ML fundamentals, LLMs, RAG, and deployment practices. Link: https://github.com/ashishps1/learn-ai-engineering

2. AI-Engineering-Hub: Practical tutorials on RAG, vector databases, and AI agent architectures. Link: https://github.com/patchy631/ai-engineering-hub

3. EdgeAI-For-Beginners (Microsoft): Learn to deploy AI models on edge devices ideal for DevOps or IoT professionals. Link: https://github.com/microsoft/edgeai-for-beginners

4. 500+ AI/ML/DL/CV/NLP Projects (Ashish Patel 🇮🇳): A massive collection of 500+ open-source projects to build and showcase your portfolio. Link: https://github.com/ashishpatel26/500-AI-Machine-learning-Deep-learning-Computer-vision-NLP-Projects-with-code

𝐀𝐝𝐯𝐚𝐧𝐜𝐞𝐝 𝐋𝐞𝐯𝐞𝐥 𝐃𝐞𝐞𝐩 𝐃𝐢𝐯𝐞 & 𝐏𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐨𝐧-𝐆𝐫𝐚𝐝𝐞 𝐒𝐤𝐢𝐥𝐥𝐬

1. NN-Zero-To-Hero: Learn deep learning from scratch code neural networks manually to master the fundamentals. Link: https://github.com/karpathy/nn-zero-to-hero

2. AIE-Book (Chip Huyen): Practical guide to real-world AI engineering and building scalable AI systems. Link: https://github.com/chiphuyen/aie-book

3. Awesome-AI-Repositories: Curated collection of advanced open-source AI engineering and infrastructure projects. Link: https://github.com/sydverma123/awesome-ai-repositories

4. Awesome-AI-Data-GitHub-Repos (Youssef Hosni): Directory of top AI, ML, NLP, and CV repositories for research and exploration. Link: https://github.com/youssefHosni/Awesome-AI-Data-GitHub-Repos

2

How to learn AI Engineering in 2025 🚀

A few people have asked me about to go about learning AI Engineering. I’m by no means an expert but I have been following the ecosystem closely, using various AI tools and trying out the various AI technologies.

I did not follow any roadmap for my own learning, but if I were to start from scratch, here’s my recommendation:

1. Technical foundation: Know programming and the basic terms of AI and ML

Be familiar with a mainstream language (Python and JS preferred). You also need software engineering skills to be able to manipulate data, call APIs, and structure applications. Have a basic understanding of key ML terms like training, inference, embeddings, fine-tuning, and evaluation.

1. Use AI tools a lot – ChatGPT, Claude Code, Copilot

Get comfortable using these tools not just for answering questions, but for coding assistance, data analysis, and brainstorming. Using these tools on a daily basis gives you a better understanding of the strengths and gaps of current state of gen AI.

1. Learn prompt engineering and techniques

Learn how to structure prompts to produce consistent, high-quality outputs by providing clear instructions, examples, and constraints. Highly recommend https://www.promptingguide.ai as well as the official guides by the frontier model makers.

1. Learn the AI engineering toolkit

Familiarize yourself with the core AI tools: LLM APIs (OpenAI, Anthropic), embedding models (Cohere, Voyage), vector storage (pgvector, Pinecone, Weaviate), and orchestration frameworks (LangChain, LlamaIndex). For example, combining OpenAI’s GPT‑4 for reasoning with pgvector for document retrieval is a standard RAG (Retrieval-Augmented Generation) setup.

1. Understand how LLM applications work under the hood

Know how multi-step flows like (RAG) fetch and re-rank information, and how agents chain tasks together. For example, a travel assistant app might retrieve relevant flight data, ask an LLM to summarize the options, then use another LLM call to generate a booking confirmation email.

1. Apply knowledge and build real projects

Apply your knowledge in hands-on projects to bridge the gap between theory and production. Build clones of famous AI apps, or common examples like chat clients, AI-powered knowledge base that answers company FAQs, meeting transcript summarizers, or a writing app with built-in tone/length controls.