Welcome to your 100-day journey toward becoming a proficient AI/ML engineer. Each day is packed with focused learning objectives, hands-on challenges, and curated resources. Follow along in order or pick and choose topics to suit your pace.
- 100 Days of Code: AI/ML Engineer Roadmap 🎓
- 📋 Contents
- Roadmap 🛣️ Overview
- Rules 📏 & Guidelines 🦮
- Phase Overview & Table of Contents
- Phase 1: Foundations 🧱 (Days 0️⃣0️⃣1️⃣ – 0️⃣1️⃣0️⃣)
- Phase 2: Hands-On LLM Book (Days 0️⃣1️⃣1️⃣ – 0️⃣3️⃣0️⃣)
- Phase 3: Prompt Engineering and LLM Augmentation (Days 0️⃣3️⃣1️⃣ – 0️⃣4️⃣0️⃣)
- Phase 4: Agentic AI (Days 0️⃣4️⃣1️⃣ – 0️⃣5️⃣0️⃣)
- Phase 5: Made-With-ML Course (Days 0️⃣5️⃣1️⃣ – 0️⃣6️⃣0️⃣)
- Phase 6: Prompt Engineering Deep Dive (Days 0️⃣6️⃣1️⃣ – 0️⃣7️⃣0️⃣)
- Phase 7: Tools, Optimization and AI Research (Days 0️⃣7️⃣1️⃣ – 0️⃣8️⃣0️⃣)
- Phase 8: Capstone Project (Days 0️⃣8️⃣1️⃣ – 1️⃣0️⃣0️⃣)
This roadmap breaks down 100 days of learning into eight progressive phases:
- Phases group related topics together.
- Daily entries include mission, why it matters, toolbox, further reading, and challenges.
- You can fork or copy each day’s section into individual README files as you see fit.
- Feel free to adjust pacing—complete multiple days in one sitting or spread them out.
- Consistency over intensity. Better to learn 1 hour/day for 100 days than cram in one weekend.
- Hands-on practice. After each concept, code an example or mini-project.
- Log ✍️ your progress. Keep a daily journal or GitHub issue to reflect on learnings.
- Share 🛜 & collaborate 🤼. Post questions or findings on Discord/Slack with
#100DaysAI. - Cite your sources. Link tutorials, papers, or videos you use for reference.
- Iterate 🔁🌀 & revisit. If a concept wasn’t clear, tag it for a deeper review later.
- Respect licensing 🫡. Use open-source code responsibly and attribute where required.
- Have fun! Mix in creative experiments — this is your journey.
Core skills in Python 🐍, math 🔢, data handling, and visualization 📉📊.
Deep dive into Large Language Models via the Illustrated LLM Book’s Colab notebooks.
Understand the basics of the prompt engineering paradigm. Figure out how to augment LLMs with techniques like retrieval-augmented generation (RAG). Understand the different RAG design patterns and implement them with vector databases.
Understand the agentic AI paradigm and build your first AI agents. Learn about the different agentic design patterns. Inspect and run 40+ example agents in categories from beginner to advanced.
End-to-end MLOps pipeline: data, training, tuning, serving, and CI/CD with MLflow and Ray. Learn about web scraping, data sourcing and data annotation for text 🔡, images 🖼️ and audio 🎶🎵. Leverage tools like Selorax, Requests and BeautifulSoup for web 🌐 scraping. Leverage tools like LabelStudio 🎙️ and CVAT for data annotation. Train ML models with PyTorch. Deploy ML models with LitServe and MLFlow.
Master prompt design, zero/few-shot, chain-of-thought ⛓️💥💭(CoT), and function calling. Prompt optimization and evaluation with tools like Opik.
Automated evaluation, model merging, fine-tuning frameworks, quantization, and advanced decoding. RAG system and LLM evaluation with tools like Opik and RAGAS. Reading and understanding AI research papers 📄📰.
Design, build, test, and deploy a full AI/ML system—your 100-day masterpiece.
Ready to embark on Day 1? Let’s get started—and don’t forget to celebrate every milestone along the way! 🚀