> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.vellum.ai/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.vellum.ai/_mcp/server.

# Welcome to Vellum

> Discover Vellum for AI Product Development. Get resources and support for prompt engineering, semantic search, agent orchestration, and more.

## Quick Links

#### [Product Documentation](/product/getting-started/overview)

Learn about Vellum's features, including Prompts, Workflows, and Deployments.

#### [Developer Documentation](/developers/getting-started/overview)

Integrate with Vellum using our SDKs, APIs, and developer tools.

#### [Agent Builder Guide](/product/agent-builder/agent-builder-sme)

Practical tips and best practices for building Workflows with Agent Builder.

#### [Workflows](/product/workflows/introduction)

Build complex AI applications with Vellum's visual Workflow builder.

#### [Evaluation](/product/evaluation/quantitative-evaluation)

Test and evaluate your AI applications with metrics and test suites.

## What is Vellum?

Vellum is an end-to-end AI development platform that helps teams build and deploy AI-powered applications. It provides a collaborative environment where both technical and non-technical team members can contribute to AI projects using familiar tools and interfaces.

![Vellum Four Pillars of Development](https://storage.googleapis.com/vellum-public/help-docs/home/vellum_four_pillars.png)

## The Challenges of AI Development

AI development is often slowed down by silos between technical and business teams, leading to:

* Slow development cycles
* Communication gaps between stakeholders
* AI systems being shipped with "vibe checks" and not rigorous evals
* Insufficient monitoring of production use cases
* Integration challenges between different tools

Vellum addresses these problems by providing a unified platform where all team members play a crucial role in the AI
development process.

## How Does Vellum Work?

Vellum combines several key capabilities into a single, integrated platform:

### For Technical Teams

* **Developer Tools**: Full IDE support and API integrations
* **Version Control**: Built-in systems for managing prompts, orchestration logic and models
* **Testing Framework**: Quantitative testing and performance monitoring. Clear visualization of inputs and outputs at each step of your graphs
* **RAG Pipeline**: Document search and retrieval-augmented generation capabilities

### For Non-Technical Teams

* **Low-Code Interface**: Visual tools for prompt engineering and testing
* **Spreadsheet Integration**: Familiar interfaces for content management
* **Collaborative Features**: Shared workspaces and feedback tools

### For Mixed Teams

There are many ways teams can collaborate in Vellum. Here are just a few common patterns:

* Subject matter experts write prompts, measure quality, and deploy updates
* Engineers build integrations and test control flow
* Changes sync between code and UI to easily share assets and feedback

This flexible approach lets each team member contribute using their preferred tools while maintaining a single source of truth.

## Getting Started

Choose your path to get started with Vellum:

#### [Developer Guide](/developers/getting-started/overview)

Get started with our SDKs, APIs, and developer tools to build and deploy AI applications.

#### [Subject Matter Expert Guide](/product/getting-started/overview)

Learn how to build AI products using our low-code interface, visual tools, and collaborative features.

Need help? Contact our support team at [support@vellum.ai](mailto:support@vellum.ai)