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

# How Rippit works

> Shared company data, your own context, and the outside world in one place. Rippit analyzes all of it at scale and turns the result into tables, reports, and agents that keep working.

Rippit has four parts. Each one works on any kind of data, structured or unstructured, from any team.

## Bring your data together

Your company's data lands in Rippit as tables. Everyone in the workspace works from the same tables, so product, marketing, sales, success, and support see the same customers, deals, and accounts.

| What it is | Examples | How it gets in |
| - | - | - |
| Sales calls and meetings | Gong, Attention, Granola, and Fireflies calls and notes | [Data integrations](https://docs.rippit.com/plg/data-integrations.md) |
| CRM and business records | Salesforce accounts, opportunities, and other objects; account, order, subscription, and usage tables | [Salesforce](https://docs.rippit.com/plg/data-integrations/salesforce.md), [Snowflake](https://docs.rippit.com/api-reference/connect/snowflake/overview.md), [Bulk API tables](https://docs.rippit.com/api-reference/connect/bulk/tables.md) |
| Internal discussion | The Slack channels you choose | [Slack](https://docs.rippit.com/plg/data-integrations/slack.md) |
| Customer conversations | Zendesk, Intercom, Freshdesk, Gorgias, Kustomer, and Pylon conversations; Decagon, Ada, and Sierra AI agent conversations | [Data integrations](https://docs.rippit.com/plg/supported-integrations.md) |
| Anything else | Chats from your own product, an internal system, a CSV export | [Bulk API](https://docs.rippit.com/api-reference/connect/bulk/overview.md), [CLI upload](https://docs.rippit.com/cli.md#upload-a-csv) |

Every source becomes a table you can query, join to other tables, and build on. Rippit reads every record in full, not a sample, a summary, or a tag.

## Add your own context

Shared data is the starting point. Each person brings what their work needs on top of it:

* **Files.** Drop images, PDFs, text files, and CSVs into a chat, or upload a CSV as a worksheet.
* **Your own connections.** Connect an MCP server for yourself only, such as your own Notion workspace, and the Rippit agent uses it in your chats. See [Who can use a connection](https://docs.rippit.com/agentic-chat/overview.md#who-can-use-a-connection).
* **Your own work.** Your chats, worksheets, and charts stay in the workspace, so you can return to them and build on them.

## Add context from outside

Some answers aren't in your data at all. Rippit reaches past it in two ways:

* **Live connections.** Through [MCP servers](https://docs.rippit.com/agentic-chat/overview.md), the Rippit agent reads and writes in the tools your teams already use: Notion, Google Drive, Glean, and Guru for company knowledge; HubSpot and Attio for CRM; PostHog and Fullstory for product behavior; Linear, Jira, and GitHub for engineering work; Profound for AI search visibility; Clay for company and contact data.
* **External research.** [Web search, people search, and agent search](https://docs.rippit.com/pipelines/research.md) bring in news, competitor pages, the right contacts at an account, and how ChatGPT, Claude, and Gemini answer a question.

## Analyze at scale

Ask a question in plain language and Rippit picks the data, reads every record it needs, and answers with the work attached. See [Ask Rippit questions](https://docs.rippit.com/plg/chat/asking-questions.md).

When the answer you need isn't a column yet, Rippit creates it. An **enrichment** applies a judgment you describe to every row: a deal-loss reason on every sales call, a competitor mention on every meeting, a product area on every request. The result is a column you can filter, count, and chart, and enrichments defined together are filled in one model call per record.

## Build pipelines

When work is worth repeating, a [pipeline](https://docs.rippit.com/pipelines/overview.md) keeps it running. It enriches every new record as it arrives, looks up related records, researches the web, rolls results up on a schedule, runs an agent on each result, and acts:

* posts to Slack or sends an email
* opens or updates an issue in Linear or Jira
* updates a CRM record
* writes to Notion or Google Drive
* calls any tool you enabled on a connected MCP server

Every step writes to a table, so each output stays in Rippit for the next question, the next pipeline, or the next person.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.