🏛️ Digital Palace

Your AI Knowledge Hub
From concepts to production in minutes, not months

Cut through AI complexity. Get to results faster.

🌟 Why Digital Palace?

• Instant AI Knowledge: Go from zero to production in minutes, not months
• Curated, Actionable Content: No fluff—just what you need, when you need it
• Always Up-to-Date: See what's new (June 2025: OpenAI o3, Claude 4, Computer Use Agents)

---

🚀 Start Here — Choose Your Path

🔥 What's Hot Right Now

• OpenAI o3 & Claude 4 - Revolutionary reasoning models

• Computer Use Agents - AI that controls your screen

• Voice AI Breakthrough - Real-time conversation

• AI Legal Compliance - EU AI Act & GDPR guide

• Today I Learned - Daily AI discoveries & insights

• SkyReels-V2: Infinite-Length Video Generation - SOTA open-source autoregressive diffusion model for long-form, high-quality video generation

• The State of AI Agents (June 2025) – Lance Martin - In-depth analysis of ambient agents, agent UX, tools, and training

• Vibe Coding & Benchmark (April 2025) – Lance Martin - How context and retrieval methods shape agent coding performance

• Gemini Embedding (Google, 2025) – Now GA - State-of-the-art multilingual embedding model for search, RAG, and more. See concept →

• Simplify your Agent "vibe building" flow with ADK and Gemini CLI (Google Developer Blog, July 2025) - Learn rapid agent prototyping with the revamped llms-full.txt file and iterative "vibe coding" workflow. See concept →

🌐 Curated Blogs & Recommended Reading

Stay ahead with the best in AI — handpicked, technical, and practical.

A practical, research-backed guide to writing clear, effective, and beginner-friendly software tutorials. Covers structure, code examples, copy-pasteability, and more—essential reading for anyone creating technical guides or documentation.

Explore these high-quality blog posts, hands-on guides, and deep dives from the AI community:

A comprehensive, free resource for mastering probability and statistics in data science. Includes a preprint textbook, 100+ Python notebooks with real-world datasets, 100+ instructional videos, and solutions to 200 exercises. Authored by Carlos Fernandez-Granda (NYU), this site is ideal for learners and practitioners seeking practical, foundational knowledge in statistics, machine learning, and data science. All materials are open-access and regularly updated.

• Agents Towards Production (Nir Diamant, GitHub)
  The open-source playbook for turning GenAI agents into real-world products. Features hands-on, code-first tutorials for every layer of production-grade agent stacks: orchestration, memory, security, monitoring, deployment, multi-agent coordination, and more. Includes ready-to-run notebooks, practical guides, and proven patterns for scaling agents from prototype to production. Ideal for developers, researchers, and teams building robust, scalable AI agent systems. (8.7k+ stars, active development)

• AI Engineering Hub (patchy631, GitHub)
  A comprehensive, hands-on repository for mastering AI engineering in production. Features in-depth tutorials on LLMs, RAGs, agentic workflows, and real-world AI agent applications. Includes ready-to-run code, practical guides, and scalable patterns for deploying advanced agents, memory, and retrieval systems. Ideal for developers, researchers, and teams building robust, production-grade AI solutions. (14.4k+ stars, active development)

• Annotated Deep Learning Paper Implementations (labml.ai, GitHub)
  A comprehensive, actively maintained collection of 60+ PyTorch implementations of influential deep learning papers. Each implementation is paired with side-by-side notes and explanations, covering transformers, GANs, RL, optimizers, normalization, and more. Ideal for researchers, students, and practitioners seeking clear, annotated code and practical insights into state-of-the-art neural network architectures. (62k+ stars, weekly updates)

• Kiro: Next-Gen Agentic IDE
  Kiro is a next-generation agentic IDE that helps you go from prompt to production with spec-driven workflows. It unpacks requirements from a single prompt, generates technical designs, creates and sequences tasks, and automates code quality with event-driven hooks. Features deep integration with your codebase, Model Context Protocol (MCP) support, and a modern AI coding experience. Docs | Download

• Context Kills VRAM: How to Run LLMs on consumer GPUs
  A practical guide to optimizing context size and memory usage for running large language models on consumer-grade GPUs. Covers real-world benchmarks, trade-offs, and actionable tips for maximizing VRAM efficiency.

• MLX-LM-LENS: Interpreting MLX Language Models (Goekdeniz Gülmez)
  Research tool for inspecting hidden states, attention scores, and embeddings in MLX-based language models. Inspired by TransformerLens, designed for Apple Silicon and MLX-LM. Useful for model interpretability and research workflows.

• Training a Rust 1.5B Coder LM with Reinforcement Learning (GRPO)
  An in-depth look at building a 1.5B parameter code LLM in Rust, using reinforcement learning and the GRPO algorithm. Explains the training pipeline, challenges, and lessons learned for open-source code models.

• Getafix: How Facebook tools learn to fix bugs automatically
  Facebook's Getafix system uses machine learning to suggest and apply bug fixes at scale. This post details the approach, real-world impact, and how AI is transforming software maintenance.

• A Visual Guide to Quantization
  A clear, illustrated walkthrough of quantization techniques for neural networks. Great for understanding how quantization reduces model size and speeds up inference, with visuals and code examples.

• PGVector on CloudSQL for GCP
  A comprehensive infrastructure-as-code solution for deploying PostgreSQL with PGVector extension on Google Cloud SQL. Features Terraform automation, multi-environment support (dev/preprod/prod), and practical examples for vector-based applications like RAG systems, semantic search, and product recommendations. Includes detailed setup guides, architecture diagrams, and best practices for production deployment.

• Block's Playbook for Designing MCP Servers
  A detailed engineering guide from Block on building robust Model Context Protocol (MCP) servers.

• 12-Factor Agents: Patterns of reliable LLM applications (Dex Horthy)
  Production-ready AI agent patterns adapted from the 12-factor app methodology. Essential for teams deploying agents at scale. Covers structured output, prompt ownership, context engineering, and production deployment patterns.

• Context Engineering: A First-Principles Handbook (davidkimai)
  A comprehensive, frontier handbook for moving beyond prompt engineering to context design, orchestration, and optimization. Inspired by Andrej Karpathy's definition of context engineering, this repository provides progressive learning from first principles to advanced neural field theory. Features hands-on tutorials, reusable templates, practical examples, and research evidence from IBM, MIT, Princeton, and other top institutions. With 2.4k stars and active development, this is the definitive resource for understanding context engineering in modern AI systems. Perfect for developers, researchers, and practitioners building production-ready AI agents.

• Hands-on Multi-Vector Retrieval with Reason-ModernColBERT in Weaviate (LightOn)
  Step-by-step notebook for advanced RAG with multi-vector embeddings and late interaction retrieval in Weaviate.

For more, see the TIL summary.

• How to Run Kimi K2 Locally (Unsloth Docs)
  Step-by-step guide to running the SOTA open-source Kimi K2 model on your own hardware. Covers quantization, hardware requirements, llama.cpp setup, prompt formatting, and real-world coding/reasoning tests. Ideal for anyone deploying advanced LLMs locally or exploring efficient model hosting.

• OpenHands: All-Hands.dev – Autonomous AI Developer Agents & Platform
  Comprehensive documentation for OpenHands, a leading open-source platform for autonomous AI developer agents. Features quick start guides for OpenHands Cloud, local installation, and advanced usage. Join the GitHub project, Slack, or Discord to connect with the community. OpenHands agents can modify code, run commands, browse the web, and automate developer workflows—ideal for anyone interested in next-generation AI coding agents and agentic automation.

• GPT-4.1 Coding Agent System Prompt (VS Code Tools Edition) – Burke Holland
  A production-grade, open-source system prompt for building advanced coding agents in VS Code. Includes detailed tool usage, todo list management, and communication guidelines for agentic workflows. Highly useful for anyone customizing Copilot or building their own AI coding agents.

• Integrating Long-Term Memory with Gemini 2.5 (Philipp Schmid)
  A hands-on guide to building chatbots with persistent, user-specific memory using Google Gemini 2.5 and the open-source Mem0 library. Covers architecture, code, and practical tips for context-aware, personalized AI assistants.

• How to build Web3 AI agents with Google Cloud (Google Cloud Blog, July 2025)
  A comprehensive, up-to-date guide to designing and deploying autonomous Web3 AI agents using Google Cloud’s Vertex AI, Agent Development Kit (ADK), and open-source frameworks (A2A, CrewAI, LangGraph). Covers agent architectures, crypto wallet integration, decentralized workflows, and production deployment patterns. Includes diagrams, code samples, and links to official tools and further resources. Essential reading for anyone building next-generation AI agents on decentralized infrastructure.

• Announcing Vertex AI Agent Engine Memory Bank (Google Cloud Blog)
  Official announcement and deep dive on Vertex AI's new managed Memory Bank service for agents. Explains how Memory Bank enables persistent, contextual, and personalized memory for conversational agents, with hands-on guides, architecture diagrams, and integration examples for ADK, LangGraph, and CrewAI. Includes links to official docs and sample notebooks. Highly recommended for anyone building production-grade, context-aware AI agents on Google Cloud.

• Topic-based Memory for Long-term Conversational Agents (arXiv, 2024)
  The foundational research behind Vertex AI Memory Bank. Proposes a novel, topic-based approach for extracting, storing, and retrieving persistent memories in conversational agents. Demonstrates how this method enables more natural, context-aware, and personalized AI interactions. Highly recommended for those interested in the science powering production agent memory.

For more author-centric and community blog links, see External Blogs.

🧩 Concepts Hub - Your Knowledge Foundation

The core of Digital Palace - Master AI concepts with intelligent cross-linking.

Why Start Here?

• 🧠 Build Understanding: 70+ concepts from basics to cutting-edge (Mental Models)
• 🔗 Smart Navigation: Every concept connects to related areas
• 🛠️ Tool Integration: Direct links to relevant tools and guides
• ⚡ Quick Reference: Fast lookup for definitions and examples

🚀 Explore the Concepts Hub →

---

🔍 How to Search This Repository

Finding what you need quickly:

1. 🔍 Use GitHub Search: Press / and search across all files
2. 📁 Browse by Category: Use the directory structure below
3. 🏷️ Filter by Tags: Look for difficulty levels (🟢🟡🔴) and categories
4. 🔗 Follow Cross-Links: Each section links to related materials
5. 📊 Check Popular Content: See most-viewed resources above

Search Tips:

• Use specific terms: "RAG system", "LangChain setup", "production deployment"
• Look for emoji indicators: 🟢 Beginner, 🟡 Intermediate, 🔴 Advanced
• Check the "Quick Access" tables in each section
• Use browser search (Ctrl+F) within any README file
• Browse Curated X Accounts for top AI voices and news

---

🚀 Getting Started

Pick Your Level

🌱 New to AI

Start Here → Try Tools → First Project
Time: 2-4 weeks → Working AI app

🔧 Can Code

Developer Path → Build Apps → Production
Time: 1-2 weeks → Production system

🧠 AI Expert

2025 Updates → Cutting-Edge Tools → Advanced Agents
Time: Ongoing → Frontier knowledge

📚 Core Resources

---

🏛️ Repository Architecture

This repository is organized as a digital palace - each section serves a specific purpose in your AI learning journey:

📁 Core Directories

digital_palace/
├── 🧩 concepts/           # Master concept index with cross-links
├── 👥 people/             # AI/ML influential figures & leaders
├── 📖 learning/           # Structured learning paths & courses
│   └── 📚 courses/        # Educational resources
├── 🎯 guides/             # Step-by-step implementation guides
│   ├── ✍️ prompting/      # Prompting techniques
│   ├── ⚡ quick-references/ # "For the Impatients" series
│   ├── 🖼️ image-generation/ # AI image guides
│   ├── 🤖 agent-development/ # AI agent SOPs
│   └── 🏋️ training/       # Training resources
├── 🛠️ tools/              # Curated tool directories & comparisons
│   └── 🧰 development-tools/ # VS Code extensions
├── 📚 reference/          # Quick lookups, APIs, cheat sheets
│   ├── 📑 technical-articles/ # Deep-dive articles
│   ├── 🧠 techniques/     # AI techniques
│   ├── 📄 research-papers/ # Academic papers
│   ├── 🗂️ datasets/       # Training datasets
│   ├── ☁️ cloud-platforms/ # Cloud guides
│   └── 🧬 genai-fundamentals/ # GenAI basics
├── 🎭 personal/           # Learning philosophy & mental models
│   ├── 💡 til/            # Today I Learned
│   ├── 🧠 mental-models/  # Decision frameworks
│   └── 💭 ideas/          # Project concepts
└── 💬 community/          # Discussions, contributions, updates
    ├── 📰 newsletters/    # Updates
    ├── 🗣️ social-content/ # LinkedIn posts
    └── 🌐 external-blogs/ # Blog recommendations

🎯 Usage Philosophy

🌱 Learn by Doing: Start with practical projects, understand theory as you build
🔄 Iterative Discovery: Return to concepts as your understanding deepens
🤝 Community Growth: Share learnings, contribute improvements, help others
📈 Continuous Updates: Stay current with the rapidly evolving AI landscape

---

🌟 What's Inside

🆕 2025 Breakthroughs

• o3 & Claude 4 - Revolutionary reasoning models
• Computer Use Agents - AI that controls your screen
• Voice AI - Real-time conversation
• AI Legal Compliance - EU AI Act & GDPR

🛠️ Tools & Frameworks

• Prompt Boost – Discover and install the best prompts, instructions, chat modes, and MCP servers for VS Code. Supercharge your agentic and AI development workflow with 1-click installation and a curated extension directory. Ideal for anyone building with AI agents, MCP, or advanced prompt engineering in VS Code.

• 158+ AI Tools - Comprehensive directory

• Framework Comparisons - LangChain vs alternatives

• Production Tools - Enterprise solutions

• Crawlee Python – Unified, asyncio-based web scraping and browser automation for Python. Ideal for data extraction, AI, and RAG pipelines. See concept page for features, install, and examples.

🎯 Practical Guides

• Zero-to-App - Build AI apps fast
• AI Agents - Autonomous systems
• RAG Systems - AI with your data
• Production Deployment - Scale to users

📚 Learning Resources

• Structured Paths - Beginner to expert
• Research Papers - Latest findings
• Mental Models - Think like an expert

---

🎯 Quick Start Guide

---

🤝 Contributing

We welcome contributions from the AI community! Here's how you can help improve this digital palace:

🌟 Ways to Contribute

• 📝 Share Knowledge: Add new articles, tutorials, or insights
• 🛠️ Tool Reviews: Submit reviews of AI tools you've used
• 🐛 Report Issues: Found broken links or outdated information?
• 💡 Suggest Improvements: Ideas for better organization or new sections
• 🔍 Fact Checking: Help keep information accurate and current

📋 Contribution Guidelines

Before submitting a tutorial or documentation, please review Rules for Writing Software Tutorials (Michael Lynch, 2025). Following these best practices ensures your contribution is clear, effective, and beginner-friendly.

1. Fork this repository
2. Create a feature branch (git checkout -b feature/amazing-contribution)
3. Make your changes following our style guide
4. Test your changes (check links, formatting, etc.)
5. Commit your changes (git commit -m 'Add amazing contribution')
6. Push to the branch (git push origin feature/amazing-contribution)
7. Open a Pull Request

🎯 High-Impact Contribution Areas

• 2025 AI Updates: Latest model releases, breakthrough papers
• Tool Comparisons: Head-to-head analysis of similar tools
• Implementation Guides: Step-by-step tutorials for specific use cases
• Performance Benchmarks: Real-world testing results
• Case Studies: Success/failure stories from actual implementations

---

📧 Connect & Support

🤝 Get In Touch

• 💼 Professional Consultation: LinkedIn - Raphaël MANSUY (If you see an error, try again later—LinkedIn sometimes rate-limits bots)
• 🐦 Latest Updates: Twitter/X - @raphaelmansuy (If you see an error, try again later—Twitter sometimes rate-limits bots)
• 📧 Direct Contact: Email

🌟 Support This Project

• ⭐ Star this repository if you find it valuable
• 🔄 Share with your network and colleagues
• 💡 Contribute your own insights and discoveries
• 📢 Spread the word about useful resources you've found here

🏢 Professional Services

Need AI implementation for your business? Raphaël offers:

• 🎯 AI Strategy Consulting - Roadmap and architecture planning
• 🚀 Implementation Support - Hands-on development and deployment
• 📚 Team Training - Upskill your developers and data teams
• 🔧 Custom Solutions - Tailored AI applications for your specific needs

---

If this repository helps you, please give it a star! ⭐

Built with ❤️ by Raphaël MANSUY for the AI community