August 2025 Summaries

The August 2025 newsletter from Temporal highlights recent updates and integrations, including the OpenAI Agents SDK pairing with Temporal for enhanced durability and the general availability of Terraform support for Temporal Cloud. It also announces a public preview of SCIM user management and namespace tags, as well as the addition of Task Queue Fairness and Nexus cancellation types in the Python SDK. The newsletter features a builder spotlight on Jason Steving's AgentOfCode project, which utilizes Temporal to solve Advent of Code puzzles, and provides links to quickstart guides for various programming languages. Additionally, it invites readers to upcoming September events like Finovate Fall and the Disney Data and Analytics Conference, where Temporal will showcase demos on workflow orchestration and scalable application development.

Multi-agent architecture, facilitated by Temporal's orchestration capabilities, simplifies the construction of durable and efficient systems where specialized agents collaborate to complete complex tasks. Multi-agent architectures are beneficial due to their ability to manage task-specific agents, reducing the complexity and evaluation challenges associated with general-purpose agents. Temporal's framework supports agent routing and task delegation, allowing for streamlined processes such as planning a trip or managing inventory issues in a fictional setting like the Harry Potter universe. The system employs a proactive approach by setting up agents to autonomously detect, analyze, and propose solutions for problems, while also incorporating human feedback through conversational agents and Model Context Protocol (MCP). This architecture model emphasizes task delegation, where each agent performs a specific role, enhancing system resilience and transparency. The design choices, such as using workflows for long-running or interactive agents and activities for short-lived tasks, align with the Detect-Analyze-Plan-Execute-Report (DAPER) pattern, enabling efficient problem-solving and execution on behalf of users.

Cargo, a startup emerging from Y Combinator's S23 batch, seeks to streamline the complex go-to-market (GTM) tech stack by acting as a central hub that connects various tools and data streams, enabling sales teams to focus on the right leads at the right time. Led by CEO Aurelien Aubert, Cargo adopts an engineering-led approach, utilizing Temporal as its core orchestration framework to manage complex, long-running workflows and ensure reliability and scalability in its operations. By leveraging a data warehouse as the central source of truth and transitioning from Postgres to ClickHouse for better handling of analytical workloads, Cargo aims to address the challenges of siloed data and disjointed customer experiences. The company's engineering mindset permeates its culture, fostering a lean team where technical understanding and autonomy are key, allowing engineers to own the development process from start to finish. Cargo's platform emphasizes smart automation over large sales development representative teams, striking a balance between robust backend infrastructure and an accessible frontend, ultimately aiming to revolutionize revenue operations by unifying disparate tools and automating complex processes.

The blog post explores how to create long-running, interactive, and reliable tools using Anthropic's Model Context Protocol (MCP) integrated with Temporal Workflows, specifically through an application developed by Temporal's Kevin Martin. This architecture, exemplified by the Temporal Invoice MCP, allows for seamless orchestration of complex, multi-step processes, enabling AI systems to handle tasks without concerns about memory or timeout issues. The post details how to interact with these workflows using tools that trigger processes, check statuses, and manage approvals, showcasing the flexibility and efficiency of integrating human-in-the-loop elements in AI systems. It emphasizes the benefits of using Temporal Workflows for durable execution, facilitating ongoing interactions with agentic systems, and highlights how tools can initiate, control, and monitor long-running processes. Additionally, readers are encouraged to explore further through community resources and a trial of Temporal Cloud.

The State of Development 2025 report reveals significant discrepancies between engineers and their managers regarding the challenges faced in software development. Based on a survey of 226 industry professionals, the report highlights the anxiety surrounding brittle systems and manual recovery processes, with engineers notably concerned about failure recovery, which does not rank highly among decision-makers' priorities. Key findings show that 94% of respondents use AI, 42% are actively building it, and 39% focus on reliability, yet only 25% of teams report smooth workflow operations, indicating a prevalence of firefighting over efficient software building. The report also sheds light on the limited influence developers have on tooling decisions in large companies and offers insights into the differing priorities between startups and enterprises. It serves as a resource for teams to benchmark their operations, advocate for better tooling, and initiate informed discussions on improving software development practices.

The text argues against using graph-based approaches for building complex, procedural logic in AI applications, particularly agentic ones, and advocates for the use of Durable Execution, which allows developers to write normal, procedural code with "crashless" guarantees. The author emphasizes that graph-based systems often fall short in handling dynamic control flows, data management, error handling, parallelism, and reusability, creating complexity and runtime failures. Durable Execution, however, supports dynamic, data-driven execution, providing incremental execution, state persistence, and fault tolerance, and allows developers to leverage the entire ecosystem of modern software engineering tools, such as source control, testing, CI/CD, and developer tooling. The article concludes that while static graphs may suit simple pipelines, durable code is far superior for dynamic, data-driven agents and cautions against the misconception that visual graphs provide a clearer or more effective representation of procedural logic.

AI agents promise transformative automation by autonomously handling complex web tasks, yet the web's human-centric design poses challenges for traditional scripts, which can break with minor changes like a button's class name alteration. Browserbase, combined with Temporal, offers a robust solution by enabling intelligent, self-healing web automation that adapts to changes and survives failures. Browserbase uses AI to semantically interact with web elements, while Temporal ensures durable execution by saving workflow states and managing retries for transient failures. This combination mitigates the issues of fragile scripts and manual fixes, allowing developers to create resilient workflows that are efficient and scalable. Browserbase and Temporal empower engineering teams to build reliable automation systems that focus on higher-value tasks, such as large-scale data collection and multi-step AI agent orchestration, by ensuring that each step is durable, retryable, and adaptable to changes.

Platform engineering teams are pivotal in creating scalable infrastructure and improving developer experiences by building internal platforms that enhance product engineers' efficiency without compromising reliability. Temporal is instrumental in supporting these teams by transforming operations into fault-tolerant, reliable workflows, thus addressing the complexities of modern applications that are often distributed and prone to issues. By automating incident response, certificate rotation, infrastructure management, CI/CD pipeline orchestration, and internal developer platform automation, Temporal helps teams move away from fragile scripts toward scalable infrastructure automation. This approach ensures operations are consistent, observable, and resilient, enabling platform teams to manage processes like service restarts, certificate renewals, resource lifecycle management, deployment workflows, and developer tasks with ease and transparency. Companies such as Datadog and Netflix have utilized Temporal to streamline operations and enhance resilience, thereby allowing engineers to concentrate on business logic over infrastructure intricacies.

Temporal has developed a system to enhance GitHub Actions' CI/CD capabilities by integrating it with their open-source orchestration framework, enabling better coordination and reliability for complex release pipelines. This integration involves using Temporal Workflows to manage the execution of GitHub Actions, providing improved error handling, observability, and the ability to handle long-running processes. The setup includes components such as GitHub App authentication, dispatch ID tracking, and Temporal Activities, which facilitate the execution and monitoring of GitHub Actions. By adopting this approach, Temporal has achieved dependable CI/CD processes with enhanced transparency, fault tolerance, and scalability, while also offering strategies for testing and production considerations. This orchestration system reflects Temporal's focus on developer productivity, providing a robust framework for managing CI/CD tasks with precision and resilience.