# PromptLayer

> PromptLayer is an AI engineering platform for managing prompts, evaluating LLM applications, and observing agents in production.

PromptLayer gives engineering and domain teams one place to version prompts, test changes against datasets, deploy approved versions, inspect production traces, and monitor quality, cost, and latency.

## Product

- [Prompt management](https://www.promptlayer.com/prompt-management/): Version, visually edit, test, release, and roll back prompts without an application redeploy.
- [LLM evaluations](https://www.promptlayer.com/evaluations/): Run dataset-backed regression tests with deterministic, human, and LLM-as-a-judge graders.
- [Agent observability](https://www.promptlayer.com/observability/): Trace multi-step agents and connect failures, cost, latency, and token usage to exact prompt versions.
- [Prompt chaining](https://www.promptlayer.com/prompt-chaining/): Build and operate multi-step LLM and agent workflows with reusable prompts and conditional logic.
- [Dataset management](https://www.promptlayer.com/dataset-management/): Collect, label, version, and reuse datasets for evaluation and fine-tuning workflows.
- [Pricing](https://www.promptlayer.com/pricing/): Compare PromptLayer plans and included platform capabilities.

## Documentation

- [PromptLayer documentation](https://docs.promptlayer.com/): Product guides and implementation documentation.
- [Quickstart](https://docs.promptlayer.com/quickstart): Add PromptLayer to an existing LLM application.
- [Prompt registry](https://docs.promptlayer.com/features/prompt-registry): Store, version, retrieve, and release prompts.
- [Evaluations](https://docs.promptlayer.com/features/evaluations): Configure datasets, evaluators, and regression testing workflows.
- [Request history](https://docs.promptlayer.com/features/request-history): Inspect logged LLM requests and metadata.

## Use cases and customer examples

- [Gorgias customer support automation](https://www.promptlayer.com/blog/gorgias-uses-promptlayer-to-automate-customer-support-at-scale/): How a support team uses prompt versioning, regression evals, and production logs.
- [Speak prompt collaboration](https://www.promptlayer.com/blog/how-speak-empowers-non-technical-teams-with-prompt-engineering-and-promptlayer/): How domain experts iterate on prompts without waiting for engineering releases.
- [NoRedInk AI grading evaluations](https://www.promptlayer.com/blog/how-noredink-used-promptlayer-evals-to-deliver-1m-trustworthy-student-grades/): How curriculum and engineering teams evaluate AI-generated grading at scale.
- [Ellipsis agent debugging](https://www.promptlayer.com/blog/how-ellipsis-uses-promptlayer-to-debug-llm-agents/): How an engineering team traces and debugs failing agent workflows.

## Comparisons and guides

- [Prompt management guides and comparisons](https://www.promptlayer.com/blog/): Articles comparing prompt management, LLM evaluation, and AI observability approaches.
- [OpenAI integration](https://www.promptlayer.com/integrations/openai/): Manage, evaluate, and observe OpenAI prompts and requests.
- [Anthropic integration](https://www.promptlayer.com/integrations/anthropic/): Manage, evaluate, and observe Claude prompts and requests.
- [LangChain integration](https://www.promptlayer.com/integrations/langchain/): Add prompt management and observability to LangChain applications.

## Company

- [About PromptLayer](https://www.promptlayer.com/)
- [Customer case studies](https://www.promptlayer.com/case-studies/)
- [Contact sales](https://www.promptlayer.com/contact/)
- [GitHub](https://github.com/MagnivOrg)