May 2026 Summaries

In 2026, the integration of AI agents in financial transactions has been enhanced by platforms like Stripe and Courier, which together facilitate a seamless human-in-the-loop approval process for AI-initiated purchases. Stripe's Agentic Commerce Suite allows AI agents to initiate purchases using a user's saved payment methods, but requires human approval before revealing payment credentials, creating a secure and accountable transaction system. Courier complements this by providing a multichannel notification system, ensuring that approval requests reach humans effectively across various platforms such as Slack, email, and SMS, with built-in escalation workflows if no response is received. This system not only improves the reliability of AI-driven financial decisions but also addresses potential risks associated with autonomous actions by requiring human oversight in critical areas like financial commitments, irreversible actions, and decisions affecting others. The robust infrastructure provided by Courier, including its Journeys API and multichannel delivery capabilities, allows for efficient communication and escalation, ensuring decisions are routed back to the AI agents promptly.

Notification infrastructure and marketing platforms serve distinct roles within a company's messaging strategy, each addressing different needs and target audiences. Notification infrastructure is event-driven, handling operational messages such as password resets or fraud alerts triggered by application code, and is typically managed by engineering and product teams. In contrast, marketing platforms focus on campaign-driven messaging like newsletters and promos, scheduled by marketing and growth teams, and involve audience segmentation and lifecycle marketing. Both systems often coexist within companies beyond the early stages, as they solve separate problems and are built on different layers of the messaging stack. Notification infrastructure is designed for real-time, operational messages with high deliverability and flexibility to switch providers without major disruptions, whereas marketing platforms are optimized for engagement metrics and generally control the delivery infrastructure themselves. The decision to use one over the other hinges on the origin of the trigger, the ownership of the message, and the reputation needs of the sender, with companies typically evolving to implement both as their communication needs diversify and grow.

Courier has streamlined the creation of customer journeys by introducing a "create journeys" API that allows multistep messaging programs to be developed using code or AI agents. This approach, termed "journeys-as-code," enables lifecycle messaging to be managed at scale, integrating seamlessly with the broader codebase and allowing for version control and iterative improvements via JSON-based journey definitions. The API supports a variety of SDKs, and the journey body consists of nodes like send, delay, branch, fetch, ai, and throttle, which are used to build complex workflows. An AI node plays a crucial role by classifying users and routing them through appropriate paths based on runtime data. This method transforms journey maintenance into a collaborative and efficient coding process, providing a more dynamic and responsive way to handle customer messaging strategies.

Courier's May 2026 update introduced significant enhancements across its platform, bridging the gap between software development and message delivery through five key launches. These updates include the general availability of the AI Agent in Journeys, which allows for tailored user experiences by classifying users and automating content generation without manual logic. The Journeys API enables the creation of customer journeys via code, facilitating integration with coding agents like Claude Code and Codex, allowing for version-controlled, programmatically generated journeys. Custom Environments provide separate stages for development, testing, and production, allowing template and journey management similar to application code. Design Studio now offers more advanced styling options, enabling brand-consistent email designs without the need for CSS overrides. Additionally, Courier Console v3 received a complete redesign for improved navigation and data visibility, enhancing user experience with features like light and dark modes and more detailed metrics. These updates collectively enhance the automation and integration capabilities of Courier, aligning messaging workflows with modern software engineering practices.

Courier has launched a comprehensive redesign of its console, known as Courier Console v3, aimed at enhancing clarity, speed, and consistency within the user interface. This overhaul includes a new visual language, modern architecture, and support for both light and dark modes, with the rollout occurring in phases over 4-6 weeks, starting with the app shell and navigation. Despite these changes, existing workflows, templates, integrations, and configurations remain unaffected, ensuring a seamless transition for users. The updated architecture facilitates faster updates and improvements, aligning with the rest of Courier's offerings. Additionally, Courier introduced a new @trycourier/courier-ui-preferences package that allows for easy integration of a notification preferences center into any web app, allowing users to manage notifications across multiple channels. Recent updates also include the release of an AI Agent in Journeys, a new Journeys API, Custom Environments, and enhanced styling controls in the Design Studio, bridging the gap between software development and message delivery.

By 2026, customer journey orchestration tools are categorized into API-first notification infrastructures, UI-first customer engagement platforms, and enterprise suites, catering to engineering, marketing, and CRM-tied organizations, respectively. These tools, including Courier, Knock, OneSignal, Novu, Customer.io, Braze, Iterable, MoEngage, Salesforce, Adobe, HubSpot, and Twilio Engage, enable tailored messaging across various channels such as email, SMS, push, and in-app notifications based on user data and events. The choice of tool depends on who manages the customer journey—engineering, marketing, or both—with engineering teams favoring API-first tools for backend integration and marketers preferring UI-first platforms for ease of use. The orchestration layer is crucial for managing message sequencing, routing, preferences, and A/B testing, sitting above basic messaging providers and email services. The tools offer features ranging from visual builders and API integrations to open-source options and are selected based on channel coverage, builder fit, data model, governance, and observability needs.

Courier has introduced a new approach to building customer engagement workflows called "Journeys-as-code," which utilizes the Journeys API as the central component. This approach allows teams to create, update, publish, and manage customer journeys programmatically using JSON, integrating seamlessly with existing customer-engagement infrastructure. The process is facilitated by multiple access points, including SDKs, the Courier CLI, and plain-English coding agents, which simplify the translation of user stories into actionable workflows. The use of coding agents, like the /courier-journeys-create skill, enables users to describe desired flows in simple terms, which are then converted into API calls, making the iteration and maintenance of customer journeys more efficient. This method addresses the issue of journeys becoming outdated or "ossified" by making updates and iterations easier and more accessible, ultimately enhancing the flexibility and responsiveness of customer engagement strategies.

The text explores the distinct differences between transactional and marketing emails, emphasizing their varying purposes, consent requirements, and infrastructure needs. Transactional emails are one-to-one, event-triggered messages necessitated by user actions, such as order confirmations or password resets, and do not require prior marketing consent. In contrast, marketing emails are campaign-driven, sent to large groups, and require explicit opt-in under regulations like GDPR and CASL. The text advises treating them as separate systems with different sending domains and IP pools to protect deliverability, as mixing them can harm transactional email performance. It also highlights the importance of authentication protocols like SPF, DKIM, and DMARC for both types of emails to enhance deliverability and compliance with regulations. Furthermore, it explains the technical infrastructure options for sending these emails, including SMTP relays, transactional email APIs, and notification orchestration layers, and underscores the importance of separating transactional and marketing streams to preserve sender reputation.

CocoaPods, a widely-used dependency manager for iOS projects, will enter a read-only state on December 2, 2026, ceasing the acceptance of new pod versions and podspecs, though existing pods will remain accessible. This shift affects a range of iOS applications, including those built with native Swift, Objective-C, React Native, and Flutter, compelling developers to transition their dependencies primarily to Swift Package Manager (SPM) or consider manual XCFramework installation for unsupported libraries. The transition is driven by the goal of reducing security risks and maintenance burdens associated with a writable public registry. While the read-only status of CocoaPods will not prevent existing applications from building, it will curtail future updates through the platform, prompting developers to adapt their build systems accordingly. To minimize disruptions, particularly for critical services like push notifications, developers are advised to inventory their Podfiles, assess dependencies for SPM compatibility, and strategically plan their migration to maintain functionality and security.

The AI node in Courier Journeys is a new feature that enhances user journeys by classifying users, branching journeys based on results, and generating personalized message content across channels from a single prompt. It simplifies the process by replacing complex conditional rules and template variants with one prompt, making journeys easier to manage and alter. The node operates by reading data from triggers, processing it through a chosen AI model, and outputting structured data that subsequent journey nodes can utilize. Setup involves defining a prompt, output schema, and model, with testing ensured before launch. The AI node is particularly valuable for scenarios requiring complex decisions and personalized communication, though it may add latency and should be used when tasks cannot be easily coded. This innovation aims to streamline personalized messaging at scale for product and growth teams without the need for extensive rule trees.

Customer journey orchestration is a sophisticated process that coordinates the messages users receive across multiple channels, such as email, SMS, push notifications, and chat, based on their interactions with a product. Unlike traditional marketing automation, which primarily serves marketers and operates through tools like Customer.io and Braze, customer journey orchestration often caters to developers and engineers, exposing functionalities as APIs, SDKs, and CLI, and allowing for integration with existing communication providers like Twilio and Slack. This orchestration layer is crucial for implementing triggers, conditional logic, multichannel sends, and retries, ensuring messages are delivered at the right time and through the most effective channel. API-first platforms, like Courier, enable detailed observability and version control as they integrate directly into source control systems, offering developers the ability to manage transactional flows, product alerts, and mixed marketing-product journeys efficiently. The system logs every step for easy debugging and optimization, helping teams improve user engagement by providing insights into delivery success rates, conversion metrics, and user drop-off points. Ultimately, choosing the right orchestration approach—whether a customer engagement platform or a developer-focused orchestration tool—depends on the team's structure and the nature of their communication needs.

Firebase's decision to end CocoaPods support for its Apple SDKs, including FirebaseMessaging, by October 2026 presents a significant transition for iOS developers, especially those relying on CocoaPods for managing dependencies in their apps. Although existing Firebase CocoaPods versions will remain functional, no new updates or fixes will be available through this platform, which could lead to outdated SDKs and potential issues with push notification delivery. Developers are advised to migrate to Swift Package Manager (SPM) or manual installation for continued updates and support. This migration involves auditing current projects for Firebase usage, updating CI/CD pipelines, and conducting thorough testing of push notification functionalities, such as APNs registration and FCM token generation, to ensure reliability. While the migration is not immediately urgent, it is a necessary step to maintain app performance and security, as Firebase Messaging plays a crucial role in the notification infrastructure of many iOS applications.

The blog post discusses the integration of AI into customer journey workflows using the AI node in Courier's Journeys platform, which allows for personalized and adaptive messaging by classifying users, scoring intent, generating notification content, and enriching user profiles based on context. AI nodes enable marketers to automate complex decision-making processes that traditionally required manual intervention, such as routing high-intent accounts for follow-up or re-engaging disengaged users. By defining structured outputs and using various AI models, including those from OpenAI and Anthropic, the platform provides real-time, context-aware recommendations and notifications, thus simplifying the creation of dynamic customer interactions at scale. The use of AI nodes allows for better personalization by leveraging prompt and schema definitions to produce consistent and actionable JSON responses, which can be tested and validated before deployment, ensuring transparency and accuracy in automated decision-making within customer journeys.