Temporal Cloud (credits)

What do you like best about the product?

What I liked most about Temporal Cloud was how easy it made managing and automating workflows. As a university student, I found it useful for working on projects without worrying too much about server management. The automatic retries and reliable workflow execution were especially helpful because they saved time and reduced errors. I also liked the clean dashboard and how everything was organized. Overall, it was a great experience and a useful tool for learning and building applications.

What do you dislike about the product?

One thing I disliked about Temporal Cloud was that it took some time to understand how everything worked, especially as a university student. The initial setup and some technical concepts were a bit confusing for beginners. I also felt that the documentation could include more simple examples and step-by-step tutorials. Other than that, my overall experience was really good.

What problems is the product solving and how is that benefiting you?

Temporal Cloud helps solve the problem of managing complex workflows and handling failures in applications. As a university student, I found it useful for my coding projects because it automatically handles retries and keeps workflows running even when something goes wrong. It saves time, reduces manual work, and lets me focus more on building and learning rather than fixing workflow issues. Overall, it makes application development easier and more reliable.

What do you like best about the product?

What I like most about Temporal Cloud is how it handles workflows that need to run reliably even when something goes wrong. As a developer, I know how tricky it can be to manage retries and keep track of long-running processes, especially when APIs are involved. I like that Temporal helps take care of that complexity instead of having to build all the retry and recovery logic from scratch. It makes the overall workflow easier to manage and lets developers focus more on the actual business logic.

What do you dislike about the product?

The main thing I dislike is that Temporal Cloud can take some time to understand at first, especially if you're new to workflow orchestration. Setting up workflows and understanding how activities, retries, and failures work requires some learning. I also feel that debugging workflows can be a bit overwhelming initially. Once you get familiar with the concepts, it makes more sense, but the learning curve could be smoother for beginners.

What problems is the product solving and how is that benefiting you?

Temporal Cloud helps solve the challenge of managing long-running workflows and dealing with failures in processes that rely on multiple API calls or services. Rather than writing custom retry and recovery logic each time something breaks, it offers a more structured way to handle these scenarios. From a developer’s perspective, I find this approach valuable because it cuts down on repetitive code and makes workflows easier to maintain over time. It also improves overall reliability when you’re working with complex integrations and many moving parts.

What do you like best about the product?

What amazed me about Temporal Cloud was how cleanly it replaced the brittle state-tracking tables and custom retry loops I used to maintain in our SQL databases. On our DPU UK & CEE Apps, I work on multi-step background processes where C# (.NET) services validate regional business rules and trigger Python workers to sync records into Qchex. Previously, if a downstream API timed out halfway through a batch, I had to manually reconcile half-completed database rows during maintenance windows. Moving those pipelines to Temporal Workflows and Activities changed everything, because the SDKs automatically persist the Event History at every step so a crashed worker or network blip simply picks up from the exact point of interruption without duplicating transactions. Personally, jumping into the Temporal Web UI timeline view and using the tcld CLI to inspect live input payloads and stack traces has saved me hours of debugging every sprint.

What do you dislike about the product?

The steepest hurdle for me and our team was getting used to Temporal's strict deterministic execution rules during Workflow Replay. When I first migrated some of our legacy C# services, I accidentally left a standard system clock call and a direct database query inside the Workflow definition instead of isolating them inside Activities, which threw non-determinism errors during testing. Setting up Worker Versioning when I roll out breaking changes to long-running .NET workflows in our CI/CD pipeline also requires careful upfront planning. Additionally, hitting the default 2MB payload limit meant I had to write a custom data converter to offload large Python batch payloads to Azure Blob Storage before passing references between steps.

What problems is the product solving and how is that benefiting you?

In a distributed enterprise architecture, keeping long-running business processes consistent across multiple microservices and databases is a massive headache. Before Temporal Cloud, our developers spent almost as much time writing defensive plumbing-dead-letter queues, exponential backoff timers, and polling jobs-as they did writing actual business logic. Temporal Cloud eliminates that boilerplate by guaranteeing durable execution out of the box while offloading the operational burden of managing Cassandra or Elasticsearch clusters ourselves. It has also become the reliability backbone for our newer AI-assisted workflows: when we build internal automation agents using Python, Claude, and MCP tool calls, wrapping each step in a Temporal Activity ensures that if an external model endpoint hits a rate limit, the workflow retries cleanly without losing context or re-running expensive prior steps.

What do you like best about the product?

I like how Temporal Cloud manages complex workflows, handles failures, and retries tasks automatically. It saves time, reduces manual work, and keeps workflows running smoothly, making our process more reliable, efficient, and easier to maintain.

What do you dislike about the product?

I found the initial setup and learning process a little challenging. The setup was fairly straightforward, but understanding the configuration took some time. Better documentation and simpler configuration options could make it easier.

What problems is the product solving and how is that benefiting you?

Temporal Cloud helps us manage complex workflows and handle failures reliably, reducing manual work. It automatically retries failed tasks, saves time, keeps workflows running smoothly, and makes our processes easier to maintain.

What do you like best about the product?

What I like best about Temporal Cloud is its reliability and the way it simplifies building and managing durable, long-running workflows. The UI is intuitive, workflow visibility makes debugging easier, and integrations with existing services and technologies help streamline development. Performance is consistent, and the managed cloud infrastructure reduces operational overhead compared with self-hosting. The documentation and developer resources make onboarding relatively straightforward. Overall, Temporal Cloud provides strong ROI by improving resilience, reducing maintenance effort, and helping teams build scalable distributed applications. Its workflow orchestration capabilities also provide a solid foundation for incorporating AI-driven processes and automating complex tasks.

What do you dislike about the product?

Temporal Cloud can have a steep learning curve for beginners, and pricing, workflow debugging, and configuration complexity can become challenging as applications scale.

What problems is the product solving and how is that benefiting you?

Temporal Cloud helps manage complex, long-running distributed workflows by providing reliable execution, automatic retries, and fault recovery, reducing manual maintenance, minimizing failures, and allowing our team to focus more on building features.

What do you like best about the product?

I like that Temporal Cloud keeps my data and storage in the cloud. I value all the tools in the program. The initial setup was super easy—I love it. I’d give it a 10; it’s the best.

What do you dislike about the product?

Nothing really

What problems is the product solving and how is that benefiting you?

So far, Temporal Cloud solves everything for me.

What do you like best about the product?

While building our multi-stage claims pipeline-coordinating AssemblyAI transcripts, Tractable vision scores, Pinecone lookups, and our Fireworks AI pricing models-writing custom database status tables and cron pollers to handle failures became a nightmare. What I like best about Temporal Cloud is how easily I can orchestrate these distributed steps directly in Python using standard asyncio patterns and @workflow.defn decorators. Instead of manually writing retry loops and tracking database flags every time an external API dropped, I just write clean sequential workflow logic. When our FastAPI backend receives a claim submission, Temporal transparently handles state persistence and automatic retries in the background. It lets me build complex asynchronous pipelines that behave like normal Python functions, knowing that if any activity fails or times out, the workflow picks up right where it left off.

What do you dislike about the product?

I hit a few frustrating roadblocks while learning Temporal's strict determinism rules in Python. Because workflows replay their event history from scratch during recovery, I couldn't use regular network calls or standard time/UUID functions inside workflow definitions, and importing heavy packages like Pydantic threw sandbox errors until I explicitly wrapped them in workflow.unsafe.imports_passed_through(). Upgrading code on long-running workflows already active in production also required careful workflow.patched checks to avoid non-deterministic crashes. Additionally, because of the 2 MB payload limit, I had to write custom storage handlers in our cache-manager to cache raw inventory spreadsheets and pass lightweight reference IDs through activity arguments instead.

What problems is the product solving and how is that benefiting you?

Temporal Cloud solved my recurring issue with "zombie workflows"-where a temporary network glitch or third-party API timeout would leave a multi-step claim stuck halfway through. For example, when our pricing workers query external search and retailer inventory APIs that rely on multiple backend engines, an intermittent 503 Service Unavailable error or engine failure used to crash our ingestion queue and leave claims hanging indefinitely. Before I implemented Temporal, I had to manually query logs to find the stuck slice and write custom scripts to re-run it. Now, every activity has its own declarative retry policy with exponential backoff, so if an external API drops or returns a 503, Temporal automatically backs off and retries transparently without manual intervention. Plus, using the managed cloud service means I don't have to spend time maintaining a self-hosted Cassandra and Elasticsearch cluster just to keep my workflow engine running.

What do you like best about the product?

We use Temporal at Sagepilot for our customer journey engine. What I like most is being able to keep the journey logic in code, with workflow state, waits and retries handled in the same model. A journey isn't always a quick background job; it can wait for a customer event before moving ahead. Temporal is a good fit for that. Having execution history to inspect also makes it easier to understand where a journey is. With Cloud, we don't have to run the Temporal server infrastructure ourselves, though we still own our workers and application logic.

What do you dislike about the product?

The learning curve is the main downside. You have to think about deterministic workflow code, activity retries and how changes affect workflows that are already running. For event-driven journeys, signal handling and workflow lifecycle also need care; Cloud doesn't remove that application-level complexity. The execution history is useful, but understanding a long workflow still takes some digging. I'd like clearer guidance around these everyday debugging and deployment cases. It's a good fit for stateful workflows, but I wouldn't use it for every small background task.

What problems is the product solving and how is that benefiting you?

At Sagepilot, our customer journeys need to coordinate multiple steps over time and react to incoming events. The problem is keeping track of where each journey is, what it's waiting for and what should happen when a step needs another attempt. Temporal gives us a durable workflow model for that, instead of having to build all the orchestration around separate jobs and status updates. The benefit is mainly on the engineering side: the flow is easier to reason about and inspect. Cloud also takes running the Temporal service off our plate. We still need to design retries, idempotency and event handling properly.

What do you like best about the product?

Temporal Cloud provides an exceptionally reliable state management layer for executing long-running incident response playbooks and automated threat remediation workflows. Unlike traditional linear SOAR tools that can fail mid-execution during network drops or API rate-limiting events, Temporal guarantees workflow state preservation and fault-tolerant execution. If a playbook step—such as querying a threat intelligence API, isolating an endpoint via EDR, or triggering an IAM revocation—fails due to transient timeouts, Temporal automatically handles exponential retries and resumes from the exact state without losing data. Writing complex security orchestration flows in standard language SDKs gives our engineering and SOC teams complete flexibility to handle complex multi-system incident response.

What do you dislike about the product?

Monitoring Action metrics and storage retention windows requires continuous oversight to keep operational costs predictable, especially when processing massive streams of low-severity SIEM events. If automated playbooks contain inefficient polling mechanisms instead of leveraging signals or async timers, Action volume can scale up rapidly. Additionally, because Temporal enforces payload limits to maintain cluster throughput, passing large raw SIEM log outputs or heavy network packet captures between workflow steps requires implementing custom object storage patterns (like AWS S3 claim-checks), which adds initial setup effort for security engineering teams.

What problems is the product solving and how is that benefiting you?

Temporal Cloud solves the issue of state loss, fragile retries, and execution failures in security automation pipelines. Prior to adopting Temporal Cloud, complex multi-step containment playbooks were prone to partial failures whenever an upstream security tool's API timed out or restarted, leaving security tickets in ambiguous states. Temporal Cloud guarantees end-to-end execution durability and provides a clear time-travel execution history in its web console, drastically cutting down time spent debugging failed security playbooks and ensuring automated threat responses complete reliably every time.

What do you like best about the product?

Client-side encryption is a proper extension point. I wrote an AES-GCM codec, ran a workflow with a test card number, then read the raw history: four payloads, all marked as encrypted, and neither the card number nor the holder's name appeared anywhere in it.

Integrations
The codec server pattern keeps decryption on our side. Our order service exposes a decode endpoint. Given an encrypted payload from the history, it returned the original card details, so authorized staff can read data without the key ever reaching Temporal.

Integrations
It fails safe. A client with the same API key but without the codec could not read the result. It stopped with an error rather than showing anything.

Integrations
The cost of encryption is modest. Across 2,000 workflows on each of three accounts, the codec added at most 5 percent of CPU time on the worker (nothing measurable in one run) and 97 bytes to each 2 KB payload. Throughput ranged from 20 percent lower to 12 percent higher than without it, which says more about the network than about the cipher. I would still measure it on our own workload before promising anyone a figure.

Performance
Pricing/ROI
The workflow page records which SDK and version ran each workflow (Python 1.34.0 in our case). During an audit that is one less thing to reconstruct from deployment logs.

UI/UX

What do you dislike about the product?

That error is accurate and unhelpful: "KeyError: 'Unknown payload encoding binary/encrypted'". A colleague without the codec sees a lookup failure in the data converter, not a message saying the payload is encrypted and which key it needs.

UI/UX
Encryption is entirely opt-in. Without a codec, payloads are stored as plain JSON, and the namespace's codec server setting is empty until somebody fills it in.

UI/UX
The quickstart produced an API key owned by the account owner. It can do everything, which is exactly why I would not hand it to a worker.

Support/Onboarding
The Agents page lists integrations for the OpenAI Agents SDK, Pydantic AI and others, but it says nothing about where the data in those agent runs goes. For a security review that is the first question.

AI/Intelligence

What problems is the product solving and how is that benefiting you?

We need durable workflows over data we may not give a third party in clear text. The codec lets us use a managed service while the encryption keys never leave our environment.