> ## Documentation Index
> Fetch the complete documentation index at: https://docs.delphina.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Welcome to Delphina

> An AI analytics agent that connects to your data warehouse and answers questions in plain English

Delphina is an AI analytics agent that connects to your data warehouse and answers questions in plain English. Ask a question, and get back SQL, numbers, charts, and written analysis — grounded in a [context layer](/context-layer) that builds itself, validates itself, and flags what it doesn't know.

## Get started

1. Sign in at [analytics.delphina.ai](https://analytics.delphina.ai) (an admin must have [connected a warehouse](/administration/warehouse-connections)).
2. Click **New Chat** and ask a question in plain English.

## The foundation: a context layer that builds itself

Delphina captures context from your warehouse schemas, query history, and analyst knowledge to create a compiled, canonical knowledge base that agents can trust. Every answer is grounded in shared definitions rather than guesswork.

<CardGroup cols={2}>
  <Card title="Ingest" icon="database">
    Reads your warehouse data and query history to build a living map of the data landscape.
  </Card>

  <Card title="Refine" icon="pen-to-square">
    Systematizes business definitions, metrics, and unwritten rules from your top analysts.
  </Card>

  <Card title="Validate" icon="check-double">
    Auto-generates evaluations for metric accuracy. Every answer passes through a built-in [critic agent](/analytics-agent/critic) with SQL lineage and org-wide observability.
  </Card>

  <Card title="Evolve" icon="arrows-rotate">
    Autonomously learns and adapts as priorities shift, schemas update, and teams ask new questions.
  </Card>
</CardGroup>

[Learn more about the context layer](/context-layer) | [Set up automated onboarding](/knowledge-agent/onboarding)

## The products your context layer powers

<CardGroup cols={3}>
  <Card title="Analytics Agent" href="/analytics-agent/overview" icon="chart-line">
    Natural-language answers with SQL, tables, charts, and analysis — reviewed by a critic agent in real time.
  </Card>

  <Card title="Data Apps" href="/analytics-agent/data-apps" icon="grid-2">
    Interactive dashboards from prompts with context-aware, refreshable KPIs and forecasts.
  </Card>

  <Card title="Workflows" href="/analytics-agent/workflows" icon="wand-magic-sparkles">
    Scheduled analyses and KPI digests that deliver narrative findings via Slack.
  </Card>

  <Card title="Research Agent" href="/analytics-agent/research" icon="telescope">
    Expert-level investigations that iterate on plans, evaluate hypotheses, and produce written reports.
  </Card>

  <Card title="Proactive Agents" href="/analytics-agent/workflows#proactive-agents" icon="bell">
    Always-on agents that flag anomalies, milestones, and insights tied to your interests.
  </Card>

  <Card title="MCP Server" href="/integrations/delphina-mcp" icon="plug">
    Point Claude Code, Cursor, or your own agents at Delphina for inherited context and answer quality.
  </Card>
</CardGroup>

## Where to use Delphina

**Web application** — full analysis, visualizations, knowledge management, and workflows at [analytics.delphina.ai](https://analytics.delphina.ai).

**Slack** — tag `@Delphina` in any channel. [Set up Slack](/analytics-agent/slack).

**MCP clients** — call Delphina from Claude Code, Cursor, or your own agents. [Set up MCP](/integrations/delphina-mcp).