---
title: Getting Started
description: Monitor the important pages AI systems reach, reuse as sources, skip, or send traffic from.
audience: [human, agent]
type: guide
version: v1
stability: stable
lastVerified: "2026-08-01"
prerequisites: []
related: [/docs/evidence-model/, /docs/mcp/, /docs/api/]
artifacts: []
---

## What is SeeLLM?

SeeLLM is the page-level evidence and action layer for AI search visibility. It combines server-side crawler visits, sampled citation checks, quoted-text evidence, and detectable AI referrals for pricing, comparison, product, docs, and top content pages.

Each signal has limits. Read the [Evidence Model](/docs/evidence-model/) before treating a fetch as a citation or a referral as a complete traffic count.

## Choose a task

Start with a bounded task rather than a broad “AI visibility” claim. The quickstarts use the same evidence model and current product contracts as the reference material.

### For people

- [Evaluate important pages](/docs/people/evaluate/)
- [Install server-side observation](/docs/people/install/)
- [Interpret evidence before acting](/docs/people/interpret-evidence/)
- [Run a weekly evidence review](/docs/people/weekly-workflow/)

### For agents and automation

- [Connect through MCP](/docs/agents/connect-mcp/)
- [Make a first REST API call](/docs/agents/first-rest-call/)
- [Work with SeeLLM jobs](/docs/agents/work-with-jobs/)
- [Receive signed Agent Hooks](/docs/agents/receive-agent-hooks/)

## Shared foundations

Read the [Evidence Model](/docs/evidence-model/) before treating a fetch as a citation or a referral as a complete traffic count. Then use the [AI Traffic Types](/docs/concepts/ai-traffic-types/) and [Reading Weekly Trends](/docs/guides/weekly-trends/) guides to interpret the record in context.

## Reference

- [REST API](/docs/api/)
- [Authentication](/docs/api/authentication/)
- [MCP tools](/docs/mcp/tools/)
- [Jobs API](/docs/api/jobs/)
- [Agent Hooks](/docs/api/agent-hooks/)