As 2025 begins, one of the most prominent trends is the rise of agents. These agents are LLMs equipped with the ability to utilize tools and work together to accomplish tasks assigned by users.

One challenge with agents is the time they can take to complete tasks, which may result in timeouts or execution errors. To address these issues, we could leverage Upstash Workflow. Upstash Workflow offers a set of features specifically designed to handle such scenarios effectively.

The Agents API of Upstash Workflow enables you to:

  • Integrate any tool compatible with AI SDK or LangChain.
  • Execute an individual agent or facilitate collaboration among multiple agents.
  • Reliably invoke agents without concerns about timeouts or transient errors.

To get started, you can refer to the Getting Started page. For more details about the features, you can refer to the Features page.

If you’re interested, you can also explore our rich examples that showcase how various architectures can be built using the Agents API:

Prompt Chaining

Sequential LLM calls where each output becomes the input for the next, enabling structured reasoning and step-by-step task completion.

Evaluator-optimizer

A feedback loop where LLM outputs are evaluated and refined iteratively to improve accuracy and relevance.

Parallelization

Distribute tasks across multiple LLMs and aggregate the results for efficient handling of complex or large-scale operations.

Orchestrator-workers

A central orchestrator directs multiple worker LLMs to complete subtasks and synthesize their outputs for complex operations.

Was this page helpful?

Suggest editsRaise issue