AI

Why AI is so good at translation and how it speeds shipping

AI translation has quietly reached human quality for the kind of copy most products ship. Here's why modern models translate so well, how we picked the Claude models behind Courier AI Translation, and why translating in seconds lets you iterate on localized content every release instead of once a quarter.

Your Entire Lifecycle Marketing Department, Run from Claude Fable 5

With the rollout of Claude' Fable model, one thing is becoming increasingly clear. Marketing execution (especially the long-tail work), will be done in an AI editor. In Courier, connect your agent to the MCP server or CLI, install Courier Skills, and keep a small folder of markdown context files. From there, one person with a coding agent covers the work that used to require a lifecycle marketer, an email designer, a marketing ops hire, and an engineer: building journeys, shipping templates, auditing every notification, and debugging delivery without opening a dashboard.

Human-in-the-loop for AI payment agents: building approval notifications that work

AI agents need human approval before taking consequential actions: financial commitments, irreversible changes, decisions that affect other people. This post covers how to design those checkpoints and build the notification infrastructure: multi-channel delivery, live context, escalation, and a back-and-forth question loop between reviewers and the agent.

What we shipped this month: May 2026 Edition

Courier shipped five launches in May 2026: AI Agent in Journeys (GA), the new Journeys API for code-driven flows, Custom Environments, Design Studio styling controls, and Courier Console v3. Each one closes a gap between writing software and shipping the messages that go with it.

Create a customer journey from AI coding agent

Use Courier's Journey API to create multistep customer engagement workflows from your coding agent of choice. Describe the kind of journey you'd like to create, answer a few questions, and publish to the platform.

Introducing the AI Node in Courier Journeys

The AI node is a new step inside Courier Journeys that classifies users, branches on the result, and writes per-channel copy from a single prompt. Here's what it does, how to set one up, and when it's worth using over a regular condition node.

AI Tools for Product Managers: The Modern PM Stack

The modern PM stack runs on AI at every step: Cursor and Claude Code for build, Pencil and Claude Design for prototyping, Courier for notifications and agent communication, Segment for routing product events and engagement data, PostHog for analytics and LLM evals, and a knowledge system like Notion for shared memory across humans and agents.

The AI Node in Journeys: Smarter Branching, Personalized Messages, and Live Enrichment

Courier Journeys now includes an AI node that you can drop into any customer journey. Use it to branch on logic too nuanced for if/then, generate message copy shaped by each user's context, enrich profiles with live data mid-flow, and batch activity into recurring digests. Existing deterministic journeys keep working. The AI node is additive, not a replacement, and it lets you unlock the kind of personalization that used to require a dedicated ML team. Here's what it does and how to use it.

Using Claude Design, Claude Code, and Courier AI to Create a Multichannel Onboarding Series in 30 Minutes

A walkthrough of building a four-part multichannel onboarding series (email, in-app inbox, mobile push, and Slack) in 30 minutes using Claude Code and Cowork to orchestrate, Claude Design to mock each channel, the courier-template-builder skill to translate mocks into Elemental JSON, the Courier MCP to publish templates, and the Courier CLI to test sends. Covers the strategy behind the sequence and recommendations for anyone doing the same.

Throttling notifications across product, transactional, and marketing streams

A notification throttle that drops every event over a limit works fine for marketing nurtures and fails for product notifications, because product events carry context users actually need. The fix is pairing throttling with auto-batch: overflow events feed a batch node that rolls them up into a single digest, optionally rewritten by an AI node that prioritizes and summarizes the contents. This guide covers per-stream throttle setups for transactional, product, and marketing flows, and how the Courier Journeys AI node and fetch-data node fit into the pattern.

How to Send Notifications from an AI Agent with Courier's MCP Server

AI agents handle support tickets, monitor pipelines, run onboarding flows, and sync CRM data. When they do, someone needs to know what happened. Courier's MCP server gives agents direct access to the full notification API: send across email, push, SMS, Slack, and in-app. Check user preferences before sending. Trigger batch and digest Journeys so high-volume agents don't spam users. This guide covers MCP server setup, CLI tooling, Courier Skills for your IDE, and five real-world patterns with starter prompts.

Build with AI: let your agent handle notifications end to end

Courier's Build with AI toolkit gives coding agents direct access to your notification infrastructure. Four integration points, one goal: let your agent send messages, debug deliveries, manage users, and follow notification best practices without context-switching. Works with Claude Code, Cursor, and any agent that can run shell commands or call MCP tools.

How to Use Claude Code on Mobile to Design, Test, and Ship Multichannel Notifications

A walkthrough of using Claude Code's mobile app with Courier's MCP server and CLI to design, test, and ship a multichannel product announcement from a phone. Covers the full workflow: drafting the content outline, creating notification templates using Courier's elemental format, building channel-specific variants for email, Slack, push, and SMS, customizing brand styling through the API, iterating on design with test sends, getting team approval, and publishing with custom routing rules. The whole thing took a couple hours, mostly cold start and design iterations.

Debug Delivery Issues in Your AI Editor with MCP

Between quarterly audits, delivery issues don’t wait. When emails fail or alerts fire, you need answers fast. Courier’s MCP server lets you debug notifications conversationally—find the message, trace its timeline, and inspect user data in minutes. Instead of jumping between dashboards, your AI agent pulls everything on demand, surfaces root causes, and even automates investigations. The result: faster fixes, less guesswork, and a smarter, always-on reliability workflow.

How to Build B2B Customer Journeys

Courier Journeys is an AI-native journey builder for multi-step customer messaging. Build on a visual canvas with branching, delays, live data fetching, AI nodes, and omnichannel sends. Define your payload schema so variables autocomplete throughout the flow. Branch on product events and profile data. Use AI to enrich profiles, drive branching logic, or generate personalized message copy. Throttle messages so customers aren't overwhelmed. Compare draft changes against the live version before publishing. Invoke and test from the CLI or MCP server. Scaffold growth patterns with Courier Skills.

Audit Notifications with Cursor or Claude Code

Most teams can't answer a basic question about their own product: what notifications are you sending, and are they working? This post walks through a full notification audit using Courier's CLI and MCP server from inside your coding environment. You'll inventory every template, pull delivery logs to surface failures, trace messages end to end, inspect rendered content for stale copy, map the data feeding each notification, break down delivery health by tenant, check preference coverage, and build a hit list of notifications to kill, revise, or batch. All from the command line. Takes about an afternoon.

Courier vs Customer.io: 2026 messaging platform comparison

Courier and Customer.io both send across multiple channels, but they're built for different jobs. Courier is a customer messaging platform for product teams: it turns the events and user data your app already produces into notifications across journeys, broadcasts, Inbox, preferences, and AI internationalization. Customer.io is built for marketers running profile-driven campaigns, segmentation, and A/B testing. This comparison covers pricing, workflows, in-app messaging, localization, and which one fits your use case.

Customer Journeys Then and Now

Most customer journey platforms don't use product data. They use marketing data. Open rates. Click-throughs. Maybe a segment based on what plan someone's on. That's not a journey. That's a drip campaign with extra steps. I spent years building behavior-based lifecycle programs at Yahoo. The decisioning was sophisticated. The organizational overhead to ship it was not. The infrastructure layer has finally caught up. Here's what behavior-based journeys looked like then, what most teams settled for, and what's actually possible now.

Give your AI agents access to your notification infrastructure with the Courier CLI

The rebuilt Courier CLI lets AI agents send notifications, debug delivery issues, and manage users, lists, and tenants across every channel. If your agent has shell access, it has access to Courier.

Best Email API Providers for Developers in 2026: SendGrid vs Postmark vs Mailgun vs SES vs Resend

Your email provider sticks with you longer than most technical decisions. Courier handles notification infrastructure for thousands of teams, so we went deep on the six email providers that show up most: SendGrid, Postmark, Mailgun, Amazon SES, Resend, and SMTP. This guide covers real API primitives, actual code from each provider's docs, Courier integration examples with provider overrides, and an honest read on where each developer experience holds up and where it breaks down. We also asked Claude to review every API and tell us which one it would wire up first. The answer surprised us.

I Built an AI Board Member in Cursor. Here's How.

Every month I send a board update—and every month I wish someone would tell me what’s wrong before it goes out. Investors are busy, feedback comes late, and most people soften the punch. So I built an AI board member using Cursor Rules: three markdown files, a basic project layout, and no plugins. Drop in your board deck, get an immediate review, and walk into the meeting with fewer surprises.

The Courier MCP Server Is Open Source. Here's How It Actually Works.

Courier's MCP server is open source at github.com/trycourier/courier-mcp. It connects AI coding tools like Cursor and Claude Code to your Courier account so they can send messages, manage users, and install SDKs without hallucinating API details. This post walks through the actual codebase: how 16 tool classes are registered (and how a config allowlist gates most of them), why we pull installation guides from GitHub at runtime instead of bundling them, how the DocsTools class generates live JWTs alongside setup instructions, and what the SdkContextTools class does in the repo to prevent v7/v8 SDK conflicts (even though it isn't wired into the server yet).