AI for Rails Teams Workshop
Teach your Rails developers to build AI features the way they build everything else: in plain Ruby, with tests, in production. Two practical days covering LLM integration, RAG, agents, evaluation and guardrails.
Everything a Rails developer needs to ship AI features
Most of the time is spent writing code. Attendees extend a workshop application step by step until it has a working, tested AI feature set.
LLMs for application developers
Tokens, context windows, temperature and model families — only the theory you need to make good engineering decisions.
Calling models from Rails
Service objects, prompts as versioned code, structured output and error handling with Ruby LLM libraries.
Streaming and background jobs
Streaming responses to the browser with Turbo Streams, and moving slow calls into jobs with retries.
Embeddings and RAG
Chunking, embeddings and similarity search with pgvector, then a complete retrieval-augmented answer flow.
Tool calling and agents
Exposing service objects as tools, building a simple agent loop, and an introduction to MCP servers.
Evals and guardrails
Testing non-deterministic code, evaluation suites in CI, prompt-injection defenses, privacy and cost control.
Two days, hands-on
-
Day 1 · Morning
Foundations and first integration
How language models behave, choosing a model, calling it from a Rails service object, prompts as code and structured output.
-
Day 1 · Afternoon
Production plumbing
Streaming with Turbo Streams, background jobs, retries and fallbacks, tracing every call, and writing tests for AI code.
-
Day 2 · Morning
Retrieval-augmented generation
Embeddings, pgvector, chunking strategies, hybrid search, permission-aware retrieval and answers that cite sources.
-
Day 2 · Afternoon
Agents, evals and safety
Tool calling, a simple agent, evaluation suites, guardrails, privacy and cost budgets — then a review of how to apply it all to your product.
A good fit if your team
- Builds and maintains a Ruby on Rails product and has AI features on the roadmap.
- Has experimented with model APIs and wants a solid, shared way of working.
- Prefers to own AI features in-house rather than depend on an outside vendor.
- Wants hands-on exercises with mentor support throughout, not a slide deck.
Workshop questions, answered
Who is the AI for Rails Teams workshop for?
Ruby on Rails developers who are comfortable with Rails and want to build AI features properly. No machine-learning background is needed. It works best when the whole team attends, including whoever reviews pull requests.
Do attendees need experience with machine learning or Python?
No. The workshop treats language models as an API your application calls. Everything is done in Ruby and Rails, with PostgreSQL and pgvector for retrieval.
Can the workshop use our own codebase?
Yes. The standard format uses a workshop application so everyone starts from the same place. For private sessions we can adapt the second day to a use case from your own product.
Is the workshop remote or on-site?
Either. We run it remotely or in person for your team, and tailor the examples to your domain.