The Cagent Project, The Bayesian Data Analysis Book, Getting Started with Claude Code
A weekly curated update on data science and engineering topics and resources.
This week's agenda:
Open Source of the Week - The cagent project
New learning resources - Getting started with Claude Code, Fine Tuning LLMs
Book of the week - The Bayesian Data Analysis
I share daily updates on Substack, Facebook, Telegram, WhatsApp, and Viber.
Are you interested in learning how to set up automation using GitHub Actions? If so, please check out my course on LinkedIn Learning:
Open Source of the Week
This week's focus is on a new open source project - the cagent, Docker, Inc’s new open-source command-line tool for building, orchestrating, and deploying intelligent multi-agent AI systems—enabling teams of specialized virtual agents to collaborate seamlessly under a root controller.
Project repository: https://github.com/docker/cagent
Highlights & Key Functionality:
Multi-Agent Runtime & Orchestration –
cagentallows you to define a team of agents, with a designated “root” agent delegating tasks to specialized sub-agents defined via simple YAML configurationDeclarative YAML Configuration – Agents are described in concise YAML files specifying their model, instructions, description, and tool access—then run with a single command, similar to Docker Compose
Rich Tool Ecosystem via MCP – Agents can leverage external tools and services through integration with Docker’s Model Context Protocol (MCP)—supporting transports like stdio, HTTP, and SSE—for enhanced capabilities
Multiple Interfaces & Deployment Options – Interact via CLI, TUI, or API server; agents can be shared via Docker registries, with event-driven streaming and strong security isolation built in
Multi-Model Support – Compatible with OpenAI, Anthropic, Gemini, DMR, and Docker AI Gateway models, making it flexible across providers
Experimental Status with Active Development – While actively under development and subject to breaking changes, it already offers powerful capabilities for building agentic workflows
More details are available in the project documentation.
License: Apache 2.0
New Learning Resources
Here are some new learning resources that I came across this week.
Getting Started with Claude Code
If you are looking for a resource to get started with Claude Code, the Claude Code Tutorial by Net Ninja looks like a great tutorial. This tutorial covers the topics such as:
Set up a project
Using the Claude MD file
Setting context
Tools and permissions
Slash commands
MCP servers
LLM Evaluation on a Custom Dataset with MLflow and Ollama
The following tutorial by Venelin Valkov focuses on LLM evaluation using MLflow to track different LLM performance KPIs.
Intro to Fine-Tuning Large Language Models
The following course by LunarTech provides an introduction to fine-tuning LLMs. The course covers core approaches for tuning LLMs, and it covers the following topics:
Intro to fine-tuning
Parameter Efficient Fine-Tuning
The QLoRA method
Pre-trained vs Fine-Tuned Model
Prompt Engineering vs. Fine-Tuning
Book of the Week
This week's focus is on Bayesian statistics. The Bayesian Data Analysis (3rd edition) by Prof. Andrew Gelman, Prof. John B. Carlin, Prof. Hal S. Stern, Prof. David B. Dunson, Prof. Aki Vehtari, and Prof. Donald B. Rubin. The book, as the name implies, focuses on the fundamentals of Bayesian statistics for data analysis. Here are some of the topics it covers:
Probability and inference
Single and multiple parameter models
Hierarchical models
Evaluating, comparing, and expanding models
Bayesian computation
Markov chain simulation
Regression models
Generalized linear models
Nonlinear and nonparametric models
Gaussian process models
The book is ideal for graduate students, researchers, and data scientists who are seeking a comprehensive, hands-on resource for Bayesian modeling—from foundational principles through modern computational techniques and applied data analysis workflows.
Thanks to the authors, the book has a free online version available on the book website.
If you want to support the authors or wish to have a hard copy, you can purchase the book on Amazon.
Have any questions? Please comment below!
See you next Saturday!
Thanks,
Rami