November 2025 Summaries

The November newsletter from Temporal provides updates on newly released features, highlights from the field, and upcoming events. Key updates include the general availability of Cloud User Groups and SCIM user management for Temporal Cloud, enabling more efficient user and permission management, as well as a public preview of Environment Configuration for enhanced environment management. The newsletter also introduces the availability of Workflow History Export and an OpenMetrics endpoint for comprehensive monitoring. This month's Builder Spotlight features Dr. Mohamed Amri's work on Durable Stateful Agents, emphasizing Temporal's ability to handle long-running workflows and transient failures without developer intervention. Additionally, a new tutorial series guides users in creating durable AI applications with Temporal, and community questions address best practices in deploying Worker fleets and understanding workflows compared to other orchestration tools. The newsletter invites readers to connect with Temporal at upcoming events, including the Replay '26 event in San Francisco and AWS re:Invent in Las Vegas, offering opportunities for live demos and community interaction.

Ishan Jain, a Senior Developer Experience Engineer at Grafana Labs, explains how the new Temporal Cloud integration for Grafana Cloud enhances the ability to visualize, monitor, and alert on Temporal Workflows' health. This integration simplifies the process by providing deep visibility into Executions, Task Queues, and capacity usage without requiring additional infrastructure, leveraging Grafana Cloud's comprehensive observability platform that combines metrics, logs, traces, dashboards, and alerts. The integration offers expanded metrics, prebuilt dashboards, cost control, and AI-assisted exploration features, allowing users to track service latency, monitor Task Queue backlogs, and identify bottlenecks efficiently. It works by scraping metrics from Temporal Cloud’s OpenMetrics endpoint and pushing them into Grafana Cloud, where metrics are automatically labeled for easy organization. Setting up this integration is straightforward and available to all Grafana Cloud users, significantly streamlining the management of Temporal Cloud workloads.

Artificial intelligence (AI) has transitioned from a hypothetical concept to a tangible element in daily developer tasks and customer-facing products, with many teams actively incorporating it into their work. Code generation is the most affected area, highlighting potential reliability issues in AI-impacted code paths. A survey of over 150 developers and technical leaders from enterprise-scale companies in North America and Europe revealed that while nearly half are still in exploration or prototyping stages, 38% consider AI essential or are scaling it in production. However, confidence in observing and debugging AI workflows remains low, with only 13% feeling very confident, and 62% experiencing measurable losses in time or revenue due to reliability problems. As teams prioritize reliability, compliance, automation, and debt reduction over the next 12–24 months, they are adopting patterns like guarded tool calls, idempotent side-effects, human checkpoints, deterministic plans, and cost and policy budgets to improve AI system stability. Durable Execution is becoming crucial for managing state, retries, and long-running tasks, with a significant uptake in orchestration adoption, especially within large companies. The widespread use of OpenAI Agents SDK, Google's ADK, and LangChain reflects a trend towards integrating orchestration in AI workflows, emphasizing the importance of durable, observable workflows and treating reliability as an essential product requirement.

Temporal is a robust platform for building AI agents, contrary to misconceptions that its requirement for deterministic workflows limits its use with non-deterministic AI models like large language models (LLMs). While Temporal workflows must be deterministic to ensure reliability and recoverability from failures, the activities within these workflows, such as LLM operations, can be non-deterministic. This separation allows AI agents to dynamically make decisions and adapt their paths using LLMs without compromising durability. By recording decision histories, Temporal enables AI agents to resume operations seamlessly after crashes, making it ideal for long-running tasks. Companies like OpenAI and Replit utilize Temporal to power sophisticated AI agents, demonstrating its capability to handle dynamic, unpredictable AI behavior while maintaining reliable execution.

The announcement of the Temporal Swift SDK at the ServerSide.swift conference introduces a significant advancement for Swift developers aiming to build durable and resilient distributed systems. This SDK leverages Swift's native async/await and structured concurrency features, allowing developers to write reliable, long-running workflows that can automatically handle crashes, failures, retries, and persistence. The SDK wraps the Temporal Core SDK using Swift's C interop and employs gRPC-swift for communication, featuring components such as @Workflow and @ActivityContainer macros, TemporalWorker, and TemporalClient for managing and executing workflows. It supports advanced patterns like child workflows and local activities while offering extensive interaction capabilities through signals, queries, and updates. The SDK, which includes examples and comprehensive support for Temporal's features, is designed for seamless integration into Swift applications and is under active development, with the community encouraged to contribute feedback and share their experiences.

On October 20, 2025, a major outage occurred when Amazon Web Services' us-east-1 region went offline, affecting numerous businesses and applications globally. Temporal, a company specializing in durable execution platforms, managed to maintain workflow continuity during the outage due to its multi-region and multi-cloud replication capabilities. Temporal Cloud detected the issue before it was officially reported and ensured that customer workflows continued to operate by allowing users to trigger failovers to alternate regions. FireHydrant, a customer of Temporal, successfully utilized these features to maintain service during the incident. Although the company faced some challenges, such as a dependency on the affected region that delayed auto-failovers, proactive measures by Temporal's engineers ensured that most customer Namespaces resumed operations in alternate regions. The company plans to address these issues to improve response times in future incidents and reinforce its commitment to providing robust and resilient services for developers.

Pydantic has developed a suite of tools aimed at enhancing the reliability and usability of AI applications, notably the Pydantic AI agent framework and Pydantic Logfire observability platform. These tools prioritize type safety, observability, and test coverage, making them suitable for production-grade AI applications. A key feature of Pydantic AI is its integration with Temporal, a platform that provides durable execution, ensuring that AI agents can manage failures, maintain state, and handle asynchronous interactions without losing progress. This integration allows developers to focus on core business logic while Temporal manages retry logic and state persistence. The framework's capabilities are demonstrated through a multi-agent Slack bot example that handles user queries with robust reliability. Pydantic encourages developers to participate in live coding sessions to explore these tools and contribute feedback for continuous improvement.