• RoboRuby
  • Posts
  • Ruby AI News - August 21st, 2025

Ruby AI News - August 21st, 2025

Tidewave launches a mind-blowing Rails coding agent

Author

Matt Solt
August 21, 2025

Pro tip: Due to the size of the newsletter, many email clients truncate the content. It is recommended to read on the web for the best experience.

Welcome to the 13th Edition of Ruby AI News! This editions features the launch of an in-browser AI coding agent for Rails, a masterclass on compounding engineering, exciting Ruby OpenAI updates, and much more.

Top Stories

Introducing Tidewave Web

José Valim and the Dashbit team have launched Tidewave Web, a browser-based coding agent for Rails and Phoenix that runs directly alongside your web application with full access to UI state, database, logs, and code context, eliminating the back-and-forth of describing what's on screen to current AI coding assistants. The agent features shared page context that automatically maps UI elements to corresponding controllers and templates, deep Rails framework integration allowing it to execute code within your running app, and collaborative browser testing where it builds and validates features in real-time using a point-and-click Javascript inspector. Currently in beta, Tidewave represents an exciting new direction in AI developer tooling. You can read a full breakdown of the announcement here.

To get started, you will need to install the Tidewave Rails gem, boot your application server in development, and visit the /tidewave route in the running application. From there, you will be prompted to log in with your Github account to create a Tidewave account.

Once you have logged in, you will need to add your Anthropic or Github Copilot API key, and you’ll be ready to start building! You can enter your first prompt in the coding agent on the left side of the screen. If you have a CLAUDE.md file, you would also need to copy or symlink that to an AGENTS.md file.

I gave it a generic prompt to start building a chat with PDF style AI application, and the coding agent worked through building the models, running migrations, testing, and generating the user interface. This has the potential to be a truly revolutionary coding experience, though it did struggle with Hotwire, rate limiting, and managing the context window, all issues I would expect to be resolved as the service moves out of beta. I encourage you to give it a try and check out the documentation.

And… stay tuned for my interview with José Valim!

Compounding Engineering

Kieran Klaassen continues to innovate with Ruby AI coding workflows in My AI Had Already Fixed the Code Before I Saw It. In what he has coined as “compounding engineering”, Kieran describes how Claude Code independently reviewed and fixed code before he even opened his laptop, having learned from three months of code reviews to apply consistent patterns without prompting. With compounding engineering, development systems become self-improving through memory and continuous learning. Unlike typical AI engineering that provides short-term gains through prompt-and-ship cycles, compounding engineering builds systems where every pull request teaches the AI, every bug becomes a permanent lesson, and every code review updates the defaults, making development exponentially faster over time. Through a detailed example of building a frustration detector for Cora, Every's AI email assistant, Kieran demonstrates how AI can iterate on its own prompts through test-driven development, running tests multiple times to ensure reliability, and learning from failures to improve success rates. The approach has transformed Cora's development metrics, reducing feature ship times from over a week to 1-3 days.

Kieran provides an actionable, five-step playbook for implementing compounding engineering: capture decisions in your CLAUDE.md file, turn failures into permanent upgrades, orchestrate multiple AI agents in parallel (planning, delegating, reviewing), make sure context files reflect your codebase, and verify outputs with tests and evaluations (see also: Kieran’s LLM evaluation gem Leva). Be sure to check out the companion video to the article as well.

Every, the organization behind Cora, will be hosting a one hour live expert session on Claude Code this Friday where you can join Kieran and other experts to learn more about Claude Code and compounding engineering. Katie Parrott of Every published a recap of the previous Claude Code camp with Claude Code Q&A: What Works, What Doesn’t, and What Will Save You Hours. And if you want to join Every, Cora, and Kieran to help build the future of email, they are looking for an exceptional Rails frontend engineer:

Github Recognizes the Ruby OpenAI Gem

Alex Rudall's Ruby OpenAI gem has achieved significant recognition by being selected as one of 71 projects for the GitHub Secure Open Source Fund, joining notable projects like Ollama, scikit-learn, and OpenCV in a three-week security hardening program that helps maintainers improve their security through funding, expert guidance, and actionable best practices. Concurrently, Alex released ruby-openai version 8.2.0, with major updates for GPT-5 compatibility, WebRTC support for OpenAI's new Realtime API, multi-image upload capabilities, and expanded documentation. The well deserved recognition by GitHub and the comprehensive release update demonstrates ruby-openai's evolution as both a secure and cutting-edge Ruby library for AI integration. Congratulations Alex!

Content

Announcement: Scott Werner introduced a new approach to AI-powered development in The System Inside the System, proposing that instead of using AI to generate entire applications, developers should create specialized AI systems that generate specific components within existing codebases. This philosophy is embodied in two new Ruby tools: vsm (Viable System Model), a Ruby gem for building self-contained, recursive systems that can act as AI agents, and airb, a CLI-based programming agent with context-aware code generation capabilities. Scott argues that Ruby’s dynamic nature may be what is needed for building systems that can evolve, and Ruby’s metaprogramming foundations may give “AI agents the ability to build themselves”.

Update: Ben Roesch released Raif v1.3.0 - Now with support for LLM evals, including LLM-as-judge. This update introduces evaluation capabilities for large language models, including the popular LLM-as-judge methodology where one LLM evaluates the outputs of another. Raif also got a new, expanded documentation site:

Documentation

Raif (Ruby AI Framework) - A Rails engine for adding AI-powered features to Rails applications

docs.raif.ai

Newsletter: Yatish Mehta launched TokenRuby: a new Ruby + AI newsletter, a weekly publication bridging the Ruby and AI ecosystems for Rails developers. Each issue features highlights from both worlds, interesting reads, and product recommendations. Yatish has previously contributed to the Ruby AI Newsletter, covering the Ruby AI Hackathon. The first three issues are available now, and you should sign up today for future updates:

Update: Carmine Paolino released five rapid-fire versions of RubyLLM, moving from 1.6.0 to 1.6.4. Version 1.6.0 introduced custom HTTP headers for beta features, tool halt control to skip LLM commentary, support for the latest models (Opus 4.1, GPT-5), new tool callbacks, and a documentation site redesign. Version 1.6.1 followed with critical fixes including OpenAI tool choice flexibility, automatic cleanup of orphaned tool messages after rate limits, and the ability to switch tool contexts mid-conversation. Version 1.6.2 focused on performance with corrected Gemini thinking token counting for accurate billing, removal of capability gatekeeping to let providers handle feature support directly, and multiple code optimizations via RuboCop Performance. Version 1.6.3 improved developer experience by allowing models to use their native temperature defaults instead of forcing 0.7, and removed problematic Overcommit hooks that required all provider API keys just to submit PRs. Version 1.6.4 added multimodal tool responses enabling tools to return images and files, fixed Gemini schema conversion issues, and included comprehensive new configuration documentation. Congratulations Carmine on passing two million downloads and being the number one trending Ruby repository on Github!

Article: Dieter S. shared insights from Building an AI Agent with Ruby and Rails from Scratch — What No One Tells You, detailing the challenges of creating a production-ready AI agent that processes and summarizes Medium posts. The post covers workflows, MCP, context, LLM choice, tool calling, and other considerations. Dieter emphasizes that while gems provide helpful abstractions, developers must still understand the underlying mechanics to build effective AI applications that can handle real-world document processing at scale.

Podcast: Joe Leo and Valentino Stoll published a recap of their interview with Obie Fernandez in From Black Boxes to a Multitude of Workers: Obie Fernandez on AI Architecture in Ruby. The discussion explores Obie's journey from traditional Ruby development to AI-powered applications, focusing on how he transitioned from building monolithic AI systems to creating distributed architectures with multiple specialized AI workers. The Ruby AI Podcast also announced their next two episodes, Episode 5: Evaluating LLMs with Rails and Leva Thanks to Kieran Klaassen, and Episode 6: Rails After the Robots: Chad Fowler on AI as the Next Abstraction.

Paper: Claire Randolph, Adam Michaleas, and Darrell O. Ricke published Large language models for closed-library multi-document query, test generation, and evaluation, introducing a system that leverages LLMs for document analysis within secure environments. The research addresses the challenge of using AI in proprietary settings where data cannot leave organizational boundaries, implementing a Ruby on Rails web interface that allows users to query multiple documents simultaneously using retrieval-augmented generation.

Update: Justin Bowen released ActiveAgent v0.5.0, introducing multimodal response capabilities with image and PDF analysis support, allowing agents to process visual content alongside text. The release improves configuration management by adding support for base client configurations via an active_agent.yml file and enhances the generator system with skip config and format options. Key fixes include correcting schema generation for user actions, fixing generation callbacks, updating the Anthropic provider to support their new API, and documentation additions.

Article: 0x1eef detailed how to implement Retrieval-Augmented Generation in 32 lines of code in An Introduction to RAG with llm.rb. The tutorial walks through using OpenAI’s Vector Stores API, generating embeddings from documentation, and using similarity search to provide relevant context to an LLM for accurate answers. In a companion post about Age Estimation with llm.rb, 0x1eef explores using OpenAI's vision models to estimate the age of a person from images, demonstrating multimodal prompts, image vision, and structured outputs. The library used, llm.rb, was recently updated to version 0.13.0.

Tutorial: Bobby Santiago demonstrated How to Add AI to Your Rails App: FastAPI Integration with Local LLMs, showing how to integrate Python-based AI services with Rails applications using FastAPI and Llama 3. The tutorial walks through building a microservice architecture where Rails handles the web application logic while FastAPI serves AI capabilities with local LLMs. The example features a sentiment analyzer for blog posts, with complete implementation details for Docker setup, service communication, and error handling. There is an accompanying demo repository on Github.

Update: Christopher Petersen announced an update to Lancelot: Ruby Bindings for Lance with Hybrid Search, providing Ruby developers access to the Lance columnar data format optimized for AI workloads. The lancelot gem enables hybrid search with semantic and full-text capabilities, efficient vector storage and retrieval, and a Ruby-idiomatic API for building AI-powered features like RAG systems and recommendation engines.

Article: Wilbur Suero posted Writing Maintainable AI Prompts in Rails with Promptly, presenting his Ruby gem Promptly for organizing AI prompts in Rails applications. The gem addresses the common problem of scattered prompt strings throughout codebases by providing a structured way to manage prompts using YAML files with ERB templating support.

Newsletter: Nate Berkopec shared the latest edition of Four Line Friday, featuring some good Ruby AI tips and tools. This week includes a production memory logging utility for Sidekiq, a shell script for setting up AI development environments, using Claude to create “platinum-plated” dotfiles, and running Claude Code instances in parallel with Conductor.

Update: Vicente Reig announced multimodal support in DSPy.rb v0.16, enabling developers to process both text and images with language models. The update includes documentation examples and supports OpenAI and Anthropic vision models with image analysis and structured outputs. The documentation demonstrates how to build multimodal chains for tasks like image-based generation and visual question answering. DSPy.rb is currently at version 0.19.1, so be sure to check the release notes for the latest updates.

Release: Vicente Reig also released BAML-inspired type definitions that promise 60-89% fewer token usage when compared to JSON Schemas. The gem, sorbet-baml, translates Sorbet types into BAML-inspired type definitions for more efficient AI interactions. The project aims to reduce token usage compared to traditional JSON schemas when sending type definitions to language models, addressing the problem of increasingly complex schemas that consume excessive tokens. The gem is designed as an alternative to JSON schemas.

Article: Rich Steinmetz explored workflows for leveraging multiple LLMs simultaneously in Code with LLMs in Parallel. The analysis looked at Git worktrees, Conductor, Claude Squard, and custom scripts, concluding that a custom setup may be the best possible solution right now. As part of a comment on the article, Ben Sheldon shared his Ruby configuration script for working with Git worktrees. Previous articles in Rich’s LLM coding series include Code with LLMs and a PLAN, Code with LLMs and default instructions, and Code with LLM teams.

Update: Dewayne VanHoozer released Ragdoll System v0.1.11: Multi-modal RAG with Enhanced Document Management. The update includes enhanced document management, Rails engine improvements, better logging, and a new job queue dashboard. He also announced upcoming support for image and audio processing, analytics, and multi-modal search.

Article: Abdelkader Boudih argues in The Mars Speech: Crossing the AI Event Horizon that AI has changed the parameters of success for developers. He focuses on junior developers and the skills and initiatives they have to adopt to become proactive winners in the AI age. He also introduces the concept of an LLM Requirements Definition Language, where engineers must learn how to communicate with LLMs properly instead of just “prompting”.

Opinion: Scott Werner explored how AI code generation is similar to the zero interest rate policy (ZIRP) era in Entering Technical Debt's ZIRP Era. He argues AI isn’t creating a technical debt crisis, its making infinite code generation feasible now because each advancement in AI will tackle previous technical debt. As AI refactoring costs approach zero, AI will fix past issues, and the winners will be those that embrace AI generation the most now. Scott is also opening up the beta for ActionsPerMinute, a gaming-inspired command center for working with different AI agents and tools across multiple projects in parallel.

Article: Justin Searls reflected on Scott Werner's ZIRP article in Is a Technical Debt ZIRP a Good Thing? While initially looking at a Technical Debt ZIRP Era as beneficial, he reassess after reviewing a thought provoking comment from Brian Marick on the post. Ultimately he concludes that the value of current LLM-based tools outweighs the risks, and the personal benefits have forever changed his approach to software development.

Release: Gabriel Dehan released claude_hooks, a Ruby DSL that simplifies creating Claude Code hooks by eliminating repetitive boilerplate of JSON parsing, STDIN/STDOUT handling, and error management. The library provides base classes for all hook types (UserPromptSubmit, PreToolUse, PostToolUse, etc.) with built-in input helpers, output manipulation methods, and session-specific logging that doesn't interfere with Claude's operations.

Podcast: Julián Duque of Heroku interviews Errol Schmidt from reinteractive in Heroku in the Wild: reinteractive on AI and Rapid POCs. The discussion focuses on how reinteractive leverages AI technologies and rapid prototyping methodologies to build proof-of-concepts quickly for their clients. They share insights on safely integrating AI into development workflows and the practical considerations for deploying AI-powered applications on Heroku.

Tips: Reddit user jalolapeno shared their workflow for effectively using Claude Code with Rails projects. The approach centers on creating a folder structure containing strategic planning documents: a plan.md summarizing the product vision, detailed MVP specifications, and most importantly, a technical.md file with implementation patterns that prevents Claude from generating poor code. When Claude struggles with specific libraries or gems, they create targeted guides and reference these in a CLAUDE.md file for quick context loading in new chats.

Release: Jared Reid released an MCP server for interacting with the RubyGems.org API, available as @ruby-mcp/gems-mcp on npm. This tool enables AI assistants to search for Ruby gems, retrieve detailed gem information including dependencies and versions, and access metadata from the RubyGems repository. When paired with a bundler/Gemfile specific sub-agent in Claude Code, it provides effective dependency management capabilities for AI-assisted development workflows.

Update: Sean McCleary updated his local gem documentation MCP server to provide CLI access to Ruby documentation and AI agent access via MCP. The gem allows AI assistants like Claude to access local Ruby gem documentation for the specific versions installed in your project, enabling accurate code assistance based on actual dependencies.

Podcast: Joël Quenneville and guest Jimmy Thigpen assesed AI workflows and limitations on The Bike Shed episode 469: How are we using AI? They look at AI blindspots, discuss their daily AI workflows, and pinpoint areas of improvement for AI agents in the future.

Update: David Heinemeier Hansson has added a markdown mime type and renderer to Rails. This should allow for Rails applications to easily provide context-specific markdown files and llms.txt files to agents. Mikkel Malmberg also shared on X/Twitter that his pull request to copy Rails error messages for coding agents has been merged. [h/t Jesse Waites]

Announcement: On a similar note, Vercel has published A proposal for inline LLM instructions in HTML based on llms.txt, proposing a new standard for embedding AI-specific instructions directly within HTML documents. This builds upon the existing llms.txt convention by allowing developers to include structured metadata and instructions that help Large Language Models better understand and interact with web content.

Release: Denis Sellu released the Brave Search Ruby Gem, a Ruby wrapper for Brave Search that includes AI-powered search summarization using the Brave native API. The gem provides an interface to access Brave's web, image, video, news, and local search functionality through their API.

Article: Josef Šimánek from Ruby Elders presented Our First Deep Dive into the RubyGems.org Ecosystem, a comprehensive analysis of the Ruby package ecosystem. To do so, Ruby Elders created an AI-powered tool to scan Github Actions workflow files for each gem, identifying testing gaps for current Ruby versions.

Video: Chad Fowler from Thoughtbot teams up with FrontrowMD's Chief Product Officer Yaser Mahmoud to explore whether ChatGPT Agent can effectively automate real-world QA processes in Automating QA with ChatGPT Agent. In Part 2: QA Automation with AI, Chad continues the investigation with Clarissa Borges, diving into alternative approaches including programmable options and screenshot-based methods to compare against the previous week's ChatGPT Agent workflow.

Tutorial: Melissa Pastore demonstrated How to use GitHub Copilot to level up your code reviews and pull requests. The post showcases how GitHub Copilot can streamline the code review process by generating detailed pull request summaries, explaining complex code changes, and suggesting improvements during reviews.

Article: Irina Nazarova from Evil Martians explores The Long Game: Why Rails Survived the Hype Cycle and What it Means for Your Startup. The article examines how Ruby on Rails has weathered technology trends and continues to be a viable choice for startups despite newer frameworks gaining attention. In the piece, Irina shares that Y Combinator partners recommend Rails for AI startups, and what that can teach us in the context of the AI hype cycle.

Events

Upcoming

August 21st - Meetup: Dresden.rb on August 21st in Dresden, Germany will have Hadi Al Qawas give a talk on Automatic generation of Ruby source code documentation with locally running LLMs.

August 21st - Meetup: The NWRUG (North West Ruby User Group) is hosting a meetup on August 21st in Manchester, United Kingdom and will feature a talk by Rahoul Baruah on Software Development with LLMs, explaining how to add LLMs to your apps, doing TDD, and writing code with AI.

August 26th - Meetup: The SF Ruby Meetup on August 26th at Github in San Francisco includes two Ruby AI speakers:

  • Miles Georgi: Tackling Domain Complexity with Foobara

  • Enrique Mogollán: What I've Learned Building an MCP Inspector in Ruby

August 27th - Meetup: The Ruby Ireland Meetup on August 27th at Intercom in Dublin, Ireland will host Declan McGrath for a demonstration on Radaica-DP - Boldly Build Without Any Gaps. In the talk, Declan will outline an AI-based approach to capturing and sustaining ideas in the software development cycle by leveraging modern AI tooling and communications to establish solid foundations for building software.

August 28th - Meetup: The Philippine Ruby User Group will meet on August 28th at Narra Labs in Quezon City, Philippines for discussions by Al Gray on FastMCP and other MCP gems, and Ray on fine tuning LLM integration and deployment with Rails 8 and Kamal.

September 4th - Conference: Rails World 2025 on September 4th and 5th in Amsterdam, Netherlands will have three Ruby AI presentations:

September 6th - Conference: XORuby on September 6th in Chicago will feature a talk by Scott Werner, the founder of Sublayer and ArtificialRuby, titled We Were Voyagers. We Can Voyage Again! The talk will focus on the rise of Generative AI as an opportunity to bring back the magic, art, and fun of creating with Ruby.

September 10th - Conference: Friendly.rb on September 10th and 11th in Bucharest, Romania will host three Ruby AI presentations:

September 12th - Conference: RubyConf India on September 12th and 13th in Jaipur, India will include four Ruby AI themed presentations, including:

  • Deepan Kumar: AI at Runtime: Teaching Your Ruby App to Fix Itself

  • Prathamesh Sonpatki: Teaching AI to Debug Your Rails Apps (So You Can Actually Sleep)

  • Vishwatjeetsingh Desurkar: What if… Ruby Led the AI Revolution?

  • Vlad Dyachenko: AI Best Practices: Understanding MCP’s Security Footprint and Managing Overpowered “Friendly Interns”

September 13th - Hackathon: Rails Fever is hosting a Rails + Claude AI: Build Together Challenge at CultureWorks on September 13th in Philadelphia. The collaborative build challenge will pair together developers to create a real Rails App in just a couple of hours, powered by Claude AI.

September 17th - Meetup: ArtificialRuby is hosting a meetup at Betaworks in New York City on September 17th. If you are interested in presenting at an ArtificialRuby meetup, please fill out this form and let them know.

September 18th - Conference: The EuRuKo (European Ruby Conference) on September 18th and 19th in Viana do Castelo, Portugal has announced seven Ruby AI & Hardware related speakers:

October 6th - Conference: The Rocky Mountain Ruby Conference on October 6th and 7th in Boulder, Colorado will host three Ruby AI speakers:

Open Source Updates

Code Spotlight

Samad Ahmed is back in the code spotlight with a new gem that categorizes transactions with RubyLLM. The repository contains a real world example of how to use RubyLLM to process transactions with a schema, categorize them by their descriptions, look up the General Ledger code for each category using a RubyLLM tool, and export the results to a CSV file.

GitHub - sahmed007/ruby-categorize-transactions

A practical real-world example of categorizing transactions with AI using Ruby - sahmed007/ruby-categorize-transactions

github.com/sahmed007/ruby-categorize-transactions

New Gems

Links to the RubyGems page, newest releases are first:

zirp - Powerful release notes and notification tool

clusterkit - High-performance clustering and dimensionality reduction for Ruby

deepagents - Deep agent with sub-agent spawning, todo list capabilities, and mock file system for Ruby

llm_translate - AI-powered Markdown translator

ccexport - Export and preview Claude Code conversations with syntax highlighting

langgraphrb_rails - Rails integration for the langgraph_rb gem

swagger_mcp_tool - MCP server for Swagger/OpenAPI specifications

raix-bedrock - AWS Bedrock provider for Raix.

claude_hooks - Ruby DSL for creating Claude Code hooks

poml - Ruby implementation of POML (Prompt Oriented Markup Language)

sorbet-baml - Convert Sorbet types to BAML type definitions for LLM prompting

promptly - Opinionated Rails integration for reusable AI prompt templates

aidp - A CLI for AI-driven software development, from analysis to execution

genai-rb - Gemini module for ruby

gpt - API client for GPT-5 (Responses API)

vsm - Async, recursive agent framework with capsules, tools-as-capsules, streaming tool calls, and observability

code_to_query - Natural language to safe, parameterized SQL for Rails

ruby_llm_community - Single API for GPT, Claude, Gemini, and with integrated community additions

code_healer - AI-powered code healing and self-repair system for Ruby applications

rcrewai-rails - Rails integration for RcrewAI - Build AI agent crews with database persistence and web UI

context_spook - Collects project context for AI

one-mentor-ruby - An API client for 1Mentor

rb-utcp - Universal Tool Calling Protocol

rcrewai - Build intelligent AI agent crews in Ruby - CrewAI implementation with multi-LLM support

langsmithrb_rails - Rails integration for LangSmith tracing and monitoring

langsmithrb - Client for LangSmith - LLM observability and evaluation platform

ColorPHMetric - Gem for evaluating colors based on their identifiers

attio-rails - Rails integration for the Attio API client

attio - Ruby client for the Attio API

WanderAndWant - Checks for wandering, wanting, searching, and other game states

ragie_ruby_sdk - Ragie API Ruby Gem

ruby_scientist_and_graphics - Ruby data science toolkit: lightweight DataFrame, cleaning, stats, plotting, simple ML, and project save/load

monogoto_api - Unofficial Ruby gem for Monogoto API

SelfModifiedDecisionTree - Aims to remove vulnerable functionalities like data marshalling to prevent certain types of attacks.

Communemashin - Gapetto Broad AI suites

annembed-ruby - High-performance dimensionality reduction for Ruby

elelem - A REPL for Ollama

pgai_rails - Ruby on Rails integration for TimescaleDB's pgai PostgreSQL extension

airb - CLI-based programming agent for Ruby with VSM architecture

top_secret - Filter sensitive information from free text

my_math_gem - Module for mathematical analysis, including calculus, linear algebra, statistics, data processing, ODE solutions, optimization, complex numbers, and Fourier transformations

New Open Source

Links to the Github repository:

AI Associate Investment Evaluation System - AI-powered investment analysis and deal flow management system for venture capital operations

AiMemory - Intelligent memory management system for AI chat applications with vector database support

ChainFetch - AI Powered Ethereum Intelligence API

Discourse AI Replier - Discourse plugin that automatically generates AI replies to forum topics to increase community engagement

AI Recruiting System - AI recruiting system powered by RubyLLM for intelligent resume analysis and candidate screening

NeonTitans - A cyberpunk-mecha-kaiju TTRPG powered by autonomous AI agents

RoR KIA AI - A model-agnostic RAG System to enhance AI model capabilities by integrating retrieval mechanisms

Workato Connector for Mistral AI - Integration of Mistral AI's advanced language model and document processing capabilities into your automation workflows

Expert Enigma - Local-first MCP server for Ruby repositories using Graph Neural Networks to provide coding agents with a structural understanding of code

Git Commit CLI - An intelligent Git commit message generator powered by AI

Inception Serve & Inception MCP - Provides remote desktop functionality for headless Chrome browsers using Chrome DevTools Protocol and Ferrum

Jobs & Opportunities

LEA uses AI-powered workflow automation for wealth management and is looking for a Founding Forward-Deployed Engineer. LEA extracts structured data from financial documents and automates workflows - handling everything from document organization to pushing clean data into CRMs and planning tools - so wealth management firms can build end-to-end client experiences without building parsing and workflow infrastructure. This is a leading opportunity to architect cutting-edge solutions with Ruby AI and the default Rails stack. LEA co-founder and CTO Chuck Blake recently presented at ArtificialRuby on his Ruby AI workflow using Command Center, an AI-driven documentation system.

Multiple

Onsite - Vietnam - RAKSUL Vietnam operates an online printing marketplace connecting businesses with printing partners, 4 OPEN POSITIONS, including:

  • Principle Engineer

  • Technical Lead

  • Senior Backend Engineer

  • Backend Engineer

Remote - Cloudwalk builds blockchain-based payment technology and financial services for merchants, 3 OPEN POSITIONS, including:

Remote - CompanyCam provides photo documentation software for contractors to organize job site images, 3 OPEN POSITIONS, including:

  • Ruby on Rails Developer

  • Fullstack Engineer

  • Senior Engineering Manager

Onsite - Spain / Georgia - Workato is an enterprise automation and integration platform unifying AI, apps, and workflows for businesses, 2 OPEN POSITIONS, including:

Remote - Rivo provides loyalty, rewards, and referral programs for Shopify stores, 2 OPEN POSITIONS, including:

Remote

.txt make LLMs speak the language of every application: Founding DevRel Engineer

Alignerr connects AI teams with expert data labelers through Labelbox's talent marketplace: Software Engineer, Ruby - AI Training

Calimala.ai matches tech talent with remote startup jobs using AI-powered recruiting: Full Stack Ruby on Rails Developer

Doximity is a professional network for doctors and healthcare clinicians: Software Engineer, Full Stack (Ruby/Rails)

GlossGenius is a salon and spa management software for booking, payments, and marketing: Software Engineer - All Levels

Housecall Pro provides field service management software for home service businesses: Senior Staff Ruby Software Engineer

Mindoula is a tech-enabled population health company delivering 24/7 behavioral healthcare and support services: Healthcare AI Engineer

Open Core Ventures funds and builds commercial companies around open source projects: AI Automation Developer

Profile.com builds intelligent search systems and recruiting tools powered by AI: Senior Rails Engineer

Reveal Technology is a defense tech company providing AI-powered 3D mapping and situational awareness tools: Full Stack Engineer: Command & Control Artificial Intelligence

Hybrid

San Diego - Lawmatics provides CRM and marketing automation software specifically for law firms: Full Stack Developer

Onsite

Atlanta / Toronto - PagerDuty is a digital operations platform for incident management and IT alerts: Staff Engineer – Ops Cloud Platform - Ruby on Rails

San Francisco - LevelPath is an AI-native procurement platform making enterprise purchasing delightful and efficient: Software Engineer / Ruby on Rails/ Back-end

India - Nagarro is a global digital engineering consultancy delivering AI-powered transformation: Associate Staff Engineer, Ruby on Rails

Romania - WeTravel is a booking and payment platform for multi-day group travel businesses and tour operators: AI Engineer

United Kingdom - Sage provides business software for accounting, payroll, and HR: Ruby on Rails Developer

Freelance

Remote - Upwork recruiter:

Are you an organization searching for an expert Ruby AI developer, or a Rubyist looking for your next development role with AI? Please reach out and let me know the type of opportunity you’re pursuing: [email protected]

One Last Thing

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.