- RoboRuby
- Posts
- Ruby AI News - May 30th, 2025
Ruby AI News - May 30th, 2025
The Ruby AI ecosystem continues to grow
Matt Solt
May 30, 2025
Welcome to the 7th Edition of the Ruby AI Newsletter! This edition provides an overview of the Ruby AI ecosystem, including some major updates since the last newsletter. Learn all about the Ruby AI hackathon, podcast, discord server, meetup, X/Twitter community, book, and much more!
Contents
Top Stories
In addition to this newsletter, the Ruby AI ecosystem now includes a wide array of resources, communities, and networking opportunities to get help building AI-enabled applications.
Ruby AI Hackathon
At the San Francisco Ruby Meetup this week, Kamil Nicieja announced that they will be hosting a Ruby AI Hackathon on July 19th in San Fransisco. This one-day hackathon will explore how AI can help solo developers supercharge Rails application development. I can’t wait to see what people will come up with!
Ruby AI Podcast
This week Valentino Stoll and Joe Leo kicked off The Ruby AI Podcast. The inaugural show features a discussion on phoenix.love, the top Ruby AI libraries, and the importance of properly evaluating AI tools. You can subscribe to the podcast on Buzzsprout.
Ruby AI Discord
Looking for real-time updates and discussions with other Ruby AI developers? Check out the Ruby AI Builders Discord run by Alex Rudall. Ask AI-related questions, discuss prompt engineering, and find updates on all of the major Ruby AI libraries. Other notable Ruby AI Discord servers include Sublayer, ActiveAgents.ai, and Langchain.rb.
Ruby AI Meetup
ArtificialRuby’s popular series of meetups in New York City continue to drive interest and adoption in the Ruby AI ecosystem. Make sure to check out their next event on July 16th and their YouTube page of previous talks.
Ruby AI X/Twitter Community
Want to network on social media with other Ruby AI builders? Check out the X/Twitter Ruby AI community run by Landon Gray.
Ruby AI Book
One of the best ways to get started with Ruby AI would be to pick up a copy of Patterns of Application Development Using AI by Obie Fernandez. Obie uses real world insights from building his AI startup Olympia to share practical patterns and principles for building intelligent, adaptive, and user-centric software systems that harness the potential of large language models and AI components.
Ruby AI Conference?
Ok I might be stretching a bit with this one. Irina Nazarova announced the San Francisco Ruby Conference, and while not entirely dedicated to AI and Ruby, it will feature a strong focus on AI tooling. Notable speakers include Obie Fernandez (creator of Raix and Roast) and Carmine Paolino (author of RubyLLM). Sign up for updates and make sure to block off your calendar for November 19 & 20. I expect a lot more AI talks to be announced soon.
Events
Previous
At RubyKaigi in April, Yukihiro "Matz" Matsumoto, the creator of the Ruby programming language, gave a presentation on “Programming Language for AI Age”. In the keynote presentation, he discussed how Ruby can dominate in the AI age, due to its conciseness, expressiveness, and extensible nature with DSLs.
ArtificialRuby posted the videos from their May event:
Chris Power presented "AI Critiques Your Vim-fu", showcasing the power of AI to coach you through Vim challenges and courses to level up your programming techniques.
Keith Harper talked about experiences as a designer working with AI to launch an MVP with “The future is here, and anyone can code it”.
Brian Fountain demoed the “World Premier of 1000 Notes”, using AI to generate a song from just a picture!
The Madrid Ruby User Group, Madrid.rb, hosted a roundtable on Real world usage of LLMs and AI in software development on May 29th.
Valencia.rb had a presentation and workshop by Vestimir Markov on model context protocol titled Learn MCP with Ruby: Connecting existing web apps and AI on May 29th.
Upcoming
The Baltic Ruby conference will have a talk by Chris Hasiński on Image Vector Search on June 14th.
The Brighton Ruby conference will feature a presentation by Valerie Woolard on vector data types and how they are used in Large Language models and AI. The talk on June 19th will discuss how LLM models may fall short and demonstrate an example of how to build a similarity search with Ruby.
ArtificialRuby is hosting a meetup in NYC on July 16th. These events fill up fast, so be sure to RSVP now if you want to attend. There are also still two speaker spots left for the event if you have something you would like to share!
Lucian Ghinda of Short Ruby News announced he will be speaking at Euroko in September on guiding ChatGPT and other LLMs to generate better, smarter Ruby tests.
CultivateLabs released version 1.1 of Raif, a Ruby AI framework for Rails apps that supports tasks, conversations, and agents. Raif integrates all major LLM providers including OpenRouter, includes seamless Rails integration, and has an impressive web UI:
The 1.1 update adds support for image, file, and PDF inputs, the ability to generate embeddings, and an expanded admin interface. CultivateLabs currently uses Raif in production with ARC Analysis, an application for geopolitical research and analysis.
Max Chernyak shared a tutorial on Getting Answers from a Big PDF with RubyLLM. The example demonstrates using RubyLLM and Google Gemini to get relevant PDF data from its table of contents.
Carmine Paolino was on episode 676 of the Ruby Rogues podcast discussing The Magic of RubyLLM.
Mike Subelsky built BundlerMCP, an model context protocol server that allows AI agents to query information about a Ruby project’s gems. Based on fast-mcp, the server helps coding agents spend less time searching the web and more time reading code and docs already stored locally.
Ed Izaguirre published the third part in his series on creating web applications with generative AI. In this article, The Simplest Possible AI Web App, he explores the most efficient techniques for building powerful AI applications entirely from a lean technology stack. Ed also detailed prompt engineering techniques for different LLM providers in this X/Twitter thread.
AI on Rails recorded Vibe Coding: Generative AI Tips, Tricks, & Techniques! The video takes a tool-agnostic look at AI-assisted coding fundamentals with practical and real world insights.
Julián Pinzón Eslava released stimulus-docs-mcp-server and turbo-docs-mcp-server, providing experimental model context protocol servers for documentation access.
I'm trying to not be on social media too much so when I come back, I'll make sure to bring some gifts:
Experimental MCP servers for Stimulus and Turbo documentation:
github.com/pinzonjulian/t…
@marcoroth_ you mind find this interesting!
— Julián Pinzón Eslava (@pinzonjulian)
7:22 AM • May 24, 2025
Stan Lo submitted a pull request to ruby-lsp to add experimental model context protocol support. This would add the ability to browse indexed classes, get information about methods, and retrieve class details such as documentation, ancestry, and available methods to the Ruby language server. A demo is available on this X/Twitter post.
Marco Roth added a pull request for a Stimulus MCP server to his stimulus-lsp library. It currently bootstraps a plethora of MCP tools for asking Stimulus-related questions about your application. Learn more and provide feedback on this X/Twitter post.
Thoughtbot published another video in their AI in Focus series for Rails called Using embeddings to filter spam. They also added two articles in their AI for Business series, Adoption challenges - people and Adoption challenges - legal, societal and ethical considerations.
DefMethod provided another update on phoenix.love, their service for generating a complete Rails test suite. The update combines flog, flay, and heckle along with the AI tooling to tackle the complexities of larger Rails applications.
Pete Hawkins of Rapid Ruby shared a video on Using Co-pilot agent to ship Rails features. He also shared this X/Twitter thread on Why Ruby on Rails is one of the best frameworks in 2025 due in large part to the available AI tooling. Be sure to check out his AI on Rails course if you haven’t already.
Francesco Kirchhoff shared a custom setup script to use OpenAI’s Codex with Rails 8. You can enable it by adding it from Environments > Your Environment > Edit > Advanced menu.
Obie Fernandez posted a thread on Refactoring god objects in the Roast framework to improve SOLID compliance. You can see the results of the refactoring generated by Anthropic’s Opus 4 model and Claude Code on Github.
Behold the power of @AnthropicAI's Opus 4 in Claude Code...
About 4 hours of human clock time yesterday and approximately $250 in tokens. It's hard to estimate the productivity boost, because...
#anthropic#ai#claudecode
— Obie Fernandez (@obie)
6:58 PM • May 24, 2025
Julián Duque of Heroku visited the Modern Web Podcast and shared What’s New About Heroku in 2025, discussing Heroku’s new AI Platform as a Service, model context protocol integrations, and AI agents. You can learn more about the platform and features on the Heroku AI page.
Javier Ramirez, the founder of PromptHero, launched RailsFast, a Rails starter app template built for AI and optimized for vibe coding.
✨ RailsFast - A Rails app template built for AI and optimized for vibe coding, so you can launch your startup over the weekend
It has everything you need to build a SaaS business on top of it. From users, to payments, to SEO
Now on presale 👉 railsfast.com
— rameerez (@rameerez)
2:04 AM • May 27, 2025
Open Source Updates
Code Spotlight
Abdelkader Boudih open-sourced a Rails application template for creating model context protocol (MCP) servers with robust tool execution capabilities. Example tools in the repository include retrieving dependency information from Bundler, code quality analysis with RuboCop, and static analysis of any Ruby code in the project directory. The application is built on top of Abdelkader’s spec-compliant MCP server / Rails engine ActionMCP.
New Versions
Links to release notes or changelog if available.
The official Anthropic gem is now generally available under the name anthropic. Alex Rudall’s Anthropic gem is available as ruby-anthropic.
aia 0.9.3rc1 - CLI for generative AI workflows, offering prompt management, shell and Ruby integration, interactive chat, and advanced automation.
cmdstan 0.5.0 - Bayesian inference for Ruby, powered by CmdStan.
documentrix 0.0.2 - Build and query vector databases, allowing users to store and retrieve dense vector embeddings for text strings.
gemini_craft 0.2.0 - Interface to generate content using Google's Gemini AI models with support for streaming, function calling, and advanced caching.
llm.rb 0.8.0 - Toolkit for Large Language Models with full support for chat, tool calling, audio, images, files, and JSON Schema generation.
llm-shell 0.5.0 - Command-line console that can interact with multiple Large Language Models (LLMs).
omniai 2.8.0 - Standardizes the APIs for multiple AI providers.
openai 0.5.0 - Library to access the OpenAI API.
polars-df 0.19.0 - Blazingly fast DataFrames for Ruby.
raif 1.1.0 - Rails engine that helps you add AI-powered features to your Rails apps, such as tasks, conversations, and agents.
roast-ai 0.2.3 - Library for running structured AI workflows along with many building blocks for creating and executing them.
sentiment_insights 0.2.0 - Analyze sentiment from survey responses, feedback, and other text sources.
vector_mcp 0.2.0 - Server-side tools for implementing the Model Context Protocol in Ruby applications.
New Gems
Links to RubyGems page.
There appears to be an official Model Context Protocol server coming in collaboration with Shopify. The source code is available on Github, and can be installed via RubyGems with the model_context_protocol_riccardo gem. It looks like Shopify has acquired the mcp gem name, but has not published under that name yet. Will update as more details become available.
ruby_llm-mcp - Client for the Model Context Protocol (MCP) that integrates with RubyLLM. Connect to MCP servers via SSE or stdio transports, automatically convert MCP tools into RubyLLM-compatible tools, and enable AI models to interact with external data sources and services.
bundler_mcp - MCP server that enables AI agents to query information about gems in a Ruby project's Gemfile, including source code and metadata.
mcpeasy - Easy-to-use MCP servers for Google Calendar, Google Drive, Google Meet, and Slack.
pangea-sdk - Ruby library to access the Pangea API.
nitpicker-code-review - Command-line tool that uses AI to review your staged Git changes and provide constructive feedback.
nvim-mcp-server - Implementation of the MCP server protocol for Neovim.
fastlane-plugin-ovo_gptdriver - Fastlane plugin that provides actions for interacting with GPT Driver.
smore - Expanded tooling for Raix and Roast.
foobara-agent - Accepts a list commands that can execute goal-oriented AI agents.
stemmers - Bindings for popular snowball stemming algorithms.
tiny_mcp - Make local MCP tools in Ruby and easily serve them.
launchdarkly-server-sdk-ai - LaunchDarkly SDK AI Configs integration for the Ruby server side SDK.
Jobs & Opportunities
Remote
Hybrid
Onsite
India - HackerOne - Software Engineer II (Ruby)
India - HackerOne - Software Engineer III (Ruby)
India - HackerOne - Software Engineer IV (Ruby)
India - Weekday AI - Sr. Ruby on Rails Backend Engineer
Other Announcements
Looking for a cool new device to hack on? Check out TRMNL, an e-ink companion screen intended to help you stay focused. The project uses Rails to power the setup page, and plugins can be written in Ruby. Do you have any ideas for AI-enabled plugins you could ship with this device?
That’s all for this edition! Be sure to reach out if you have any stories, content, jobs, or events you want featured in the newsletter.