Stealth logo

Stealth

Ruby framework for building text and voice chatbots

Rails-inspired Ruby framework for creating conversational bots with MVC architecture, multi-platform support, and built-in state management backed by Redis and Sidekiq.

Stealth banner

Overview

Convention Over Configuration for Conversational AI

Stealth is a Ruby framework designed for developers who want to build production-ready text and voice chatbots using familiar Rails-like conventions. It adopts an MVC architecture where views are replaced with "replies," providing a structured approach to conversational design.

Scalable, Platform-Agnostic Architecture

Built as a Rack application, Stealth deploys anywhere Ruby runs. Its variant system lets you maintain a single codebase while supporting multiple messaging platforms—Facebook Messenger, Twilio SMS, Bandwidth, Smooch, and even Alexa Skills. Sessions follow a state-machine pattern backed by Redis, while incoming webhooks process asynchronously through Sidekiq queues for horizontal scalability.

Extensible by Design

Stealth's modular architecture separates service integrations into standalone gems. Add natural language processing via Microsoft LUIS or AWS Comprehend, track engagement with Mixpanel analytics, or integrate additional messaging channels as your needs evolve. The framework includes best practices out of the box: catch-all error handling, hello flows, and goodbye flows ensure robust conversational experiences without boilerplate code.

Highlights

Rails-inspired MVC architecture with state-machine sessions
Single codebase deploys across multiple messaging platforms via variants
Asynchronous webhook processing with Sidekiq and Redis backing
Extensible gem ecosystem for NLP, analytics, and messaging channels

Pros

  • Familiar Rails conventions reduce learning curve for Ruby developers
  • Rack-based deployment works with any Ruby hosting environment
  • Built-in best practices for error handling and conversation flows
  • Modular architecture keeps core framework lightweight and focused

Considerations

  • Ruby-only ecosystem limits appeal for polyglot teams
  • Requires Redis and Sidekiq infrastructure for production use
  • Voice support currently limited to early-alpha Alexa Skills
  • Smaller community compared to JavaScript-based bot frameworks

Managed products teams compare with

When teams consider Stealth, these hosted platforms usually appear on the same shortlist.

Landbot logo

Landbot

No-code chatbot builder for conversational experiences

ManyChat logo

ManyChat

Chat marketing automation on Instagram, WhatsApp, TikTok, Messenger

Rasa logo

Rasa

Open-source conversational AI framework

Looking for a hosted option? These are the services engineering teams benchmark against before choosing open source.

Fit guide

Great for

  • Ruby teams building multi-platform conversational interfaces
  • Developers seeking Rails-like structure for chatbot projects
  • Projects requiring horizontal scalability with queue-based processing
  • Teams wanting modular NLP and analytics integrations

Not ideal when

  • Teams without Ruby expertise or existing Rails infrastructure
  • Simple bots that don't require state management or scalability
  • Projects needing mature voice assistant support beyond Alexa
  • Organizations standardized on Node.js or Python bot frameworks

How teams use it

Multi-Channel Customer Support Bot

Deploy a single conversational agent across Facebook Messenger, SMS, and Smooch while maintaining unified business logic and state management.

E-Commerce Shopping Assistant

Build a stateful bot that guides users through product discovery, cart management, and checkout using NLP for intent recognition and Redis-backed sessions.

Voice-Enabled Alexa Skill

Create an Alexa skill using the same MVC patterns and reply structures as text-based bots, sharing core logic across voice and messaging channels.

Lead Qualification Chatbot

Implement a conversational form with state-machine flows, catch-all error handling, and Mixpanel analytics to track conversion funnel performance.

Tech snapshot

Ruby100%
HTML1%

Tags

conversational-botsbot-frameworkvoicerailsconversational-aiconversational-agentsbotsalexa-skillbotconversational-uinatural-language-processingfacebook-messenger-botchatbot-frameworkchatbotrubyalexa-skills-kitstealth

Frequently asked questions

What messaging platforms does Stealth support?

Stealth supports Facebook Messenger, Twilio SMS, Bandwidth, and Smooch through separate gems. Voice support includes early-alpha Alexa Skills integration.

Do I need Rails experience to use Stealth?

While Stealth follows Rails conventions like MVC architecture, you only need Ruby knowledge. Familiarity with Rails patterns will accelerate development but isn't required.

What infrastructure does Stealth require?

Stealth needs Redis for session storage and Sidekiq for asynchronous webhook processing. As a Rack app, it deploys to any Ruby-compatible hosting environment.

Can I add custom NLP or analytics services?

Yes. Stealth's extensible architecture allows you to create custom gems for any service. Existing integrations include Microsoft LUIS, AWS Comprehend, and Mixpanel.

How does Stealth handle conversation state?

Sessions use a state-machine pattern backed by Redis, allowing you to define conversation flows with transitions, validations, and context persistence across messages.

Project at a glance

Stable
Stars
595
Watchers
595
Forks
59
LicenseMIT
Repo age8 years old
Last commit3 months ago
Primary languageRuby

Last synced 4 hours ago