Curated links for AI builders

Large language model course. From basics to RAG, agents, and fine-tuning.

Course from dlthub on building robust ELT pipelines. Includes a holiday lesson (Dec 22) on integrating LLMs into your workflow, with 50 swag packs to compete for.

Learn how to build modern, scalable data pipelines using Python and AI-assisted tools. This hands-on tutorial starts from the basics of data ingestion and takes you all the way to advanced techniques in data loading, transformation, deployment, and automation. By freeCodeCamp, featuring Alexey Grigoriev and Adrian Brudaru.

Free cohort-based version running December and January. Complete the project and review three other submissions to receive a certificate of completion signed by Alexey.

A free introductory crash course on agentic AI that explains how modern AI agents work in practice, from tools and RAG to memory, planning, MCP, and multi-agent systems. Designed as a clear, realistic starting point focused on real-world system design and limitations rather than hype.

Programming assignments for

New cohort starts on January 12, 2026. A free 9-week course on building production-ready data

Curated resource for

Materials from Tim Warner's O'Reilly Live Learning

Tutorial by Nader Dabit for building a persistent AI assistant from scratch (starting with a Telegram bot + Anthropic API) and iteratively adding sessions, memory, tool use, and scheduled tasks.

Implementation-heavy Stanford course that walks through building a language model end to

AI copilot for data science. One prompt runs the full

AI-powered app builder. Design in the browser, export to GitHub.

AI-first code editor. Built on VS Code, with Copilot-style assistance.

Command-line coding agent. Scaffold and edit projects from natural language.

Structured AI agents in Python. Type-safe, testable agent workflows.

Framework for LLM applications. Chains, agents, and integrations.

Provider-agnostic MCP server that turns your AI CLI or IDE into a coordinator for multiple

Developer-friendly NLP wrapper for LLMs. Run NER, classification, and more with minimal code and zero training data. Converts unstructured model output into reliable, structured Python objects for production.

Agent Skill that lets Claude delegate coding tasks to the OpenAI Codex CLI for multi-model collaboration. Claude coordinates and refines; Codex handles implementation, debugging, and code analysis in a sandbox.

Mobile, web, and CLI client for Claude Code and Codex. Run and monitor from anywhere with E2E encryption, push notifications when the agent needs attention, and seamless switching between desktop and phone.

Run Claude Code in an isolated, reproducible Docker environment without changing how you use the CLI. Sandboxed file access and credentials for security and reliability; supports all Claude Code options.

Lets AI agents control your Chrome browser via a lightweight extension using the full Playwright API with minimal context. Reliable browser

Claude Code plugin for native multi-agent

Open-source agent skills for data

Agentic framework that automates publication-ready methodology diagrams and statistical plots directly from paper text, references, or rough sketches—optimized for scientific accuracy and visual consistency.

Autonomous financial research agent that plans, executes, and validates analysis using real-time market data, with safety features like loop detection and step limits.

Battle-tested AI coding practices for Claude Code and Cursor to keep an effective 80/20 AI-to-review ratio using disciplined context management and intentional review rituals.

Claude Code skill that turns long, complex tasks into a file-based workflow using persistent Markdown files. Stores plans, progress, and errors on disk to reduce goal drift and repeated mistakes across many tool calls or sessions.

Workflow showing how to use the OpenAI Codex CLI for deep, non-interactive debugging via a file-based question-and-answer

Lightweight, modular framework for building AI agent pipelines emphasizing atomic, single-purpose components that are reusable, composable, and predictable (built on Instructor and Pydantic).

Orchestration system for managing multiple Claude Code instances simultaneously. Coordinates parallel tasks, tracks work across agent instances, manages merge queues, and keeps persistent agent identities (tmux-based workflow).

Open-weight reasoning

Curated list of ML resources. Frameworks, papers, and tools.

Open standard for AI coding

Curated library of real-world Claude use cases across research, writing, coding, analysis, and everyday work. Organized by role, industry, and feature with concrete, end-to-end examples.

Large open-source GitHub

Curated list of tools, patterns, and projects built around slash-command interfaces. Practical reference for command-driven workflows, bots, and developer tools.

Curated collection of AI agent use cases across healthcare, finance, education, retail, and more. Maps practical applications to open-source implementations and frameworks (CrewAI, AutoGen, Agno, LangGraph). Hands-on inspiration hub for builders and practitioners.

Practical resource that combines a live webinar recording with the full planned write-up and a demo project. It also compares the AI Engineer role to traditional data team roles and maps modern AI projects to CRISP-DM.

In-depth tutorial from Pinecone's FAISS learning series that covers LSH (theory + Python implementation) for approximate nearest-neighbor search.