Set up Rig, from sign-up to your first answer

    Rig Ingest pulls your business data into a warehouse and builds an AI-ready context layer on top. This is a guided tour of getting set up end to end, from sign-up to your first dashboard and skill, following a walkthrough by Rig's founder, Toby.

    Watch the tour

    Sign up and get verified

    Rig is approved-access while we onboard our first customers. After you sign up, a member of the Rig team reviews and approves your workspace. Once you are approved, open Connections in the left sidebar and work through the steps below. Each one shows a tick when it is complete.

    Provision a warehouse

    Your data needs somewhere to land. Choose a Rig-hosted warehouse and we provision a private one for you on MotherDuck, or Bring your own warehouse: BigQuery, Snowflake, Redshift, Databricks, Microsoft Fabric, MotherDuck or Postgres. A read-only connection to your own warehouse is enough for Rig to build context on what is already there; add a write credential as well if you want Rig to sync your sources into it. To switch later, an owner opens the Warehouse step under Connections → Connectors and uses Change warehouse. It keeps your connected sources, their schedules and credentials. Data already synced is not copied across, so re-sync your sources to fill the new warehouse. Moving off a Rig-hosted warehouse drops its database, which stays recoverable for 7 days, and the dialog asks you to type your workspace's identifier to confirm. Detach, also owner-only, is for starting over: it removes every connected source as well, so do not use it to switch.

    For technical audiences

    Rig-hosted warehouses are completely tenant isolated. Every client gets a separate warehouse that we manage, so nothing is shared across customers. Prefer to bring your own? Rig reads your warehouse, writes into it only if you give it a write credential, and leaves the data already there untouched.

    Choose a language model

    Rig uses a language model to read your warehouse, build context and answer questions. Choose Use Rig-hosted LLM or Bring your own key. You can swap providers at any time without re-syncing your data.

    Your data stays yours
    The Rig-hosted model runs with do-not-train enforced and per-workspace trace isolation, so your data is never used to train anyone's model. For keys, providers and what it costs, see LLM & Billing.

    Connect your sources

    This is where you connect the tools your data lives in: your CRM, support desk, billing, product analytics and more. Search for a source, open its card and click Connect. For most sources you enter credentials, click Continue, choose the streams you want and click Connect. Some, such as Google Drive, use a form of their own; each connector's guide walks through its exact steps. The source then appears in the list with its own row: click Sync there to load it straight away, or Test to pull a small sample first. Adding sources needs a workspace owner or a role with the Manage connections permission. Each source lands in its own schema in your warehouse, so nothing is mixed together. New sources sync nightly; to change that, or to see what a source's status means, read Keep your data fresh.

    The Rig Connections screen showing connected sources and an Add a source grid where a connector moves from Test to Syncing to a successful sync.
    Connections

    Sources

    Each connector lands in its own per-source schema. Test pulls a small probe; Sync runs the full pipeline.

    Connected
    Attio
    Attio
    synced 2d ago·success
    Sync
    HubSpot
    HubSpot
    synced 2d ago·success
    Sync
    Add a source
    Search Attio, Slack, Stripe, …
    ✓ ConnectedAttio
    Attio
    CRM
    ✓ ConnectedHubSpot
    HubSpot
    CRM
    TestStripe
    Stripe
    Billing
    Slack
    Slack
    Communication
    Meta Ads
    Meta Ads
    Advertising
    Beam
    Beam
    Marketing
    PostgreSQL
    PostgreSQL
    Database
    + 300 more

    Per-source setup guides
    Every source connects a little differently. Some use an API key, some a token, and some a one-click sign-in. Browse the full catalog of sources Rig connects on the integrations page, or jump straight to a step-by-step setup guide:Connect HubSpotConnect SalesforceSalesforce access for migration & auditConnect ZendeskConnect JiminnyConnect Beam (Lightmeter)Connect StrapiConnect ShopifyConnect TikTok ShopConnect XeroConnect Google AdsConnect Google Analytics 4Connect Google DriveConnect Google Cloud StorageConnect KlaviyoConnect GorgiasConnect QuickBooksConnect Triple WhaleConnect Amazon Seller CentralConnect Amazon AdsConnect PostscriptConnect FairingConnect NorthbeamConnect AppLovinConnect Microsoft AdsConnect Meta AdsConnect TikTok AdsConnect TrustpilotConnect ArchiveConnect Instagram BusinessConnect SupabaseConnect MixpanelConnect PendoConnect AttioConnect SlackConnect LinearConnect GranolaConnect LumaConnect JiraConnect ActiveCampaignConnect AffinityConnect Agree.comConnect AircallConnect AirspeedConnect AirtableConnect AliaConnect AnthropicConnect Claude ComplianceConnect Claude EnterpriseConnect ApifyConnect Apollo.ioConnect AsanaConnect AWINConnect AWS BedrockConnect BigQueryConnect BigQuery shared datasetsConnect BladePRO (i-fulfilment)Connect BoxConnect CalendlyConnect CharlieHRConnect CircleConnect ClickUpConnect CriteoConnect CrunchbaseConnect CSV / Parquet (S3)Connect DatabricksConnect DatadogConnect DropboxConnect Microsoft Dynamics 365Connect Microsoft Dynamics 365 Finance & OperationsConnect FathomConnect Fireflies.aiConnect Frate ReturnsConnect FreshcallerConnect FreshchatConnect FreshdeskConnect FrontConnect fulfilmentcrowdConnect FX rates (ECB)Connect Gainsight CSConnect GitHubConnect GmailConnect GoAffProConnect GongConnect Google CalendarConnect Google Search ConsoleConnect Google SheetsConnect Google Tag ManagerConnect Google Workspace DirectoryConnect GreenhouseConnect HappenstanceConnect HeyReachConnect IntelligemsConnect IntercomConnect Jungle Scout (Cobalt)Connect Jungle Scout Data CloudConnect KnoCommerceConnect LangfuseConnect LightConnect LinkedIn AdsConnect MailchimpConnect MakeConnect Marketo EngageConnect MercuryConnect Microsoft ClarityConnect Microsoft FabricConnect MintsoftConnect MixShiftConnect MomentumConnect monday.comConnect MongoDBConnect MotherDuckConnect MySQLConnect NetSuiteConnect NooksConnect NotionConnect OkendoConnect OneDriveConnect OrderspaceConnect OutlookConnect parcelLabConnect PimentoConnect Pinterest AdsConnect PipedriveConnect PostgreSQL / SupabaseConnect PostHogConnect Public file linkConnect RampConnect RedshiftConnect Reviews.ioConnect Revolut BusinessConnect RichpanelConnect RipplingConnect SCORM CloudConnect SellerboardConnect SentryConnect SFTP / FTPSConnect ShipBobConnect ShipHeroConnect ShopMyConnect SkimlinksConnect SkioConnect Snapchat AdsConnect SnowflakeConnect Snowflake metadataConnect SpecterConnect Sprout SocialConnect SQL ServerConnect Stord (OMS)Connect StoryclashConnect StripeConnect SumsubConnect ThinkificConnect TracksuitConnect TypeformConnect UnleashedConnect VeeqoConnect Walmart MarketplaceConnect WorkableConnect Workleap OfficevibeConnect YouTubeConnect ZencargoConnect Zendesk AI AgentsConnect Zuora

    Upload files

    Owners and roles with the Manage connections permission can upload files. Spreadsheets and documents go in through the upload area on the CSV / Parquet (S3) tile's form. Leave the label field above it empty, since that names a bucket connection. Once you have a bucket connection the field is marked as required, but uploading works with it empty. CSV, TSV, Parquet, JSON, JSONL, NDJSON, Excel, PDF, Word, Markdown and plain-text files are accepted. Each data file becomes its own table, and documents become rows of text that Rig can search. In a JSON file each object becomes a row, and the file must hold one JSON object per line. Each file can be up to 50 MB; for anything larger, sync it from a Google Drive folder or your own bucket. Uploads land in a source of their own, listed as file_uploads, and a sync of it starts as soon as the upload finishes, or on the next sync if one is already running. The file types are covered in more detail in Keep your data fresh.

    A tool that is not listed

    If a tool has no connector but can export on a schedule, have it export CSV or Parquet files into a Google Drive folder and connect that folder with the Google Drive source, so new files arrive without anyone uploading them. If it has no export either, tell us which tool it is.

    Build your context layer

    Once your sources are in, Rig builds the context layer automatically: table and column descriptions grounded in real sample rows, detected relationships between tables, a business glossary, grain hints and certified metrics. This is what lets your AI tools answer accurately instead of guessing. Watch progress on the Build context step, and rebuild any time you add a new source.

    Go deeper on rules and metrics
    Learn how usage rules and governed metrics work in Teach Rig your tribal knowledge.

    Build a dashboard in Claude

    In Claude, with the Rig MCP connector turned on, you can ask for an interactive dashboard. Claude pulls the data from Rig and builds it, and you can ask Claude to save it back as an automation on Rig.

    Back in Rig, that dashboard now lives in Data Apps. You can see the underlying data behind it, and you can share or embed it so colleagues with the right permissions can access it too.

    The full dashboard playbook
    Walk through building governed dashboards step by step in Vibe code dashboards on internal data with Claude Cowork.

    Create a skill

    As you work with Rig you can also create skills. For example, you can build a skill that takes your sales calls and applies Rob Snyder's pull framework to categorize each call by how strongly the customer expressed pull for what you are building. It uses a Rig workflow to build a data model from that, which you can then use in your analytics. Ask Rig to review yesterday's sales calls, categorize them by how strong the pull is, and update the CRM with that information.

    Turn a question into an automation
    See how to schedule recurring work in Automate your first workflow.

    Add the MCP to your tool

    Once you are set up and your data is ingested, open Connections and choose the MCP tab. Its MCP Setup section has the instructions for adding the Rig MCP to your tool of choice, so you can start querying your data from your own tools.

    The MCP Setup panel in Rig with seven collapsed sections: Add MCP to Claude Code, Add MCP to Cursor, Add MCP to Codex, Adding Rig as a Cowork Connector, Add to Claude (claude.ai — one click), Agent instructions (optional) and Always allow read-only tools (optional)

    Connect your AI tools
    For per-tool setup, see Connect Rig to your AI tools. Questions? Send them through the contact form.

    Was this guide helpful?