Let every team answer their own data questions
Rig sits on the Snowflake, BigQuery or Databricks you already run and gives sales, finance, CS and marketing plain-English access to it. The numbers are defined once, so everyone gets the same answer. You keep your pipelines and your models.
// In plain terms
Rig is AI-native data infrastructure. Think Stripe or Vercel, but for data.
We connect all your sources, build a layer on top of them that knows what every table means, and let your team and your AI tools reach that data safely. For a B2B SaaS company that means the warehouse stops being a thing only the data team can use, and net revenue retention, churn and CAC payback mean the same thing in every deck.
// What you'd build
What a B2B SaaS company builds first
A company at Series A or later usually has the data: the warehouse runs, the models exist, the CRM is clean enough. What is missing is a way for the people who need an answer to get one without booking analyst time.
Net revenue retention, defined once
NRR, gross churn and logo churn agreed as one definition and certified, so the number in the board pack, the number in the sales deck and the number finance reports are the same number. This is the one that usually has three answers.
Snowflake · BigQuery · Stripe · Salesforce · HubSpot
Churn early warning
Accounts with usage falling three weeks running and a renewal inside ninety days, surfaced before the renewal call. Birdie runs churn reporting and escalations through Rig.
Mixpanel · PostHog · Zendesk · HubSpot
Win/loss from the whole record
Why a deal died, read across the tickets, the calls, the product usage and the invoices rather than from the two words a rep had time to type into the closed-lost field.
Salesforce · HubSpot · Zendesk · Stripe · Jiminny
Seat usage and expansion signals
Seats provisioned against seats active in the last 30 days, by account and by plan, so expansion conversations start from what the customer is using.
Mixpanel · PostHog · Stripe · Salesforce
CAC payback by cohort
What each cohort cost to acquire and when it paid back, blended and paid-only, measured against closed pipeline rather than attributed leads.
Google Ads · Meta Ads · HubSpot · Stripe
Call coaching from transcripts
Rep calls compared against the top performers, with objection-handling patterns and coaching opportunities drawn from outcomes. Novakid coaches country leads this way.
Jiminny · Fireflies · Granola · Fathom
The board and investor pack
The pack is assembled from the source systems against the certified definitions each cycle, not rebuilt by hand from last quarter's deck with the dates changed.
Stripe · Xero · NetSuite · Salesforce
Governed access for every team and agent
Rig MCP switched on org-wide so Claude, ChatGPT, Cursor and Codex read the same governed layer a person would, each scoped to its own role and logged. Birdie went from 7 to 19 active users this way in a month.
Rig MCP · Claude · ChatGPT · Cursor · Codex
// In production
"The result is going to be consistent every single time, and now I can't imagine operating without it."
Judy, Chief Customer Officer at Birdie
Where it fits in the stack you already bought
Rig is headless by design here. Ingestion is optional, because you already have pipelines. If you run a dbt Semantic Layer or Cube project, Rig syncs those models in rather than asking you to define everything twice.
What Rig owns is the layer no other tool maintains: the context that tells an agent what a table means, which metric is the certified one, and who is allowed to see it.
// Proof
B2B SaaS companies on Rig, and the companies alongside them
Birdie, Novakid, deskbird and Vita Mojo are the SaaS deployments, each on a warehouse the company had before Rig.
// Built on Rig
What B2B SaaS teams build on Rig
Each of these starts from the CRM, product and billing data in your warehouse.
Closed-lost & win-loss analysis
See why deals really die, beyond what the rep typed
Pull the full context around every deal from CRM, support, product and finance, so you can see what actually slowed deals down and what moved win rate.
BirdiePipeline forecasts
A forecast built from deal evidence in place of stage percentages
Roll your pipeline up into a forecast that weighs each deal's real signals (activity, engagement, history), so the number you take to the board is one you can defend.
Churn-risk & renewals prep
Spot at-risk accounts before the renewal call
Score renewal risk across product usage, support tickets, CRM and finance, then auto-assemble the renewals or QBR pack so CSMs walk in prepared.
BirdieAccount health & expansion
One health score across product, support and finance
A live account health view that blends usage, support load and revenue signals to flag churn risk early and surface the accounts ready to expand.
Quote-to-cash & invoicing
Automate the path from closed-won to cash collected
A workflow that picks up every closed-won deal, raises and reconciles the invoice across CRM, billing and your ledger, and chases what is overdue.
Investor & board reporting
Auto-assemble the board pack from source data
Generate the monthly board and investor update straight from finance, product and CRM data, with the numbers and commentary built in minutes.
in minutesSelf-serve dashboards
Let anyone build a custom dashboard in plain English
Business teams ask for the report they need and Rig builds a governed, custom dashboard on your modelled data, so the data team stops being a ticket queue.
CleoCommon questions
No. Birdie runs Snowflake and Novakid runs BigQuery, and Rig sits on top of both. Rig reads your existing models and syncs a dbt Semantic Layer or Cube project in rather than asking you to rebuild it. What Rig adds is the context layer that keeps definitions current as the schema drifts, governed access for people and agents, and the hosting layer for what they build.
Commercial teams, not the data team. Birdie went from 7 to 19 active Rig MCP users month on month, across sales, finance, marketing, customer success and the exec team, and runs 15+ dashboards and agents weekly. Novakid has 10 commercial users building on Rig MCP, with country leads asking their own questions rather than queueing for an analyst.
A text-to-SQL tool guesses at your schema every time it is asked, so the same question phrased two ways returns two answers. Birdie hit exactly that before Rig, querying Notion, Snowflake, Intercom and HubSpot through general agent tools. Rig grounds every query in a governed context layer with certified metrics, and runs it in a sandbox with role-based access and an audit trail.
Yes. Salesforce, HubSpot, Attio, Pipedrive and Gainsight on the CRM side, Jiminny, Fireflies, Granola and Fathom for calls, Mixpanel and PostHog for product analytics, and Linear, Jira, GitHub and Sentry for delivery. Novakid uses call transcripts to compare a rep's calls against top managers worldwide and surface coaching opportunities.
Rig builds the context layer over your warehouse, you certify the metrics that matter to your commercial teams, then MCP goes on for the org and each function builds its own workflows. Birdie's buyers were commercial leads, not the data team, which is the usual shape here.
Access is role based, per person and per agent, so a rep sees their accounts and finance sees the ledger. Queries run in a sandbox and every one is logged. See the security page for what is in place today, including our current SOC 2 status.
Put governed access on the warehouse you run
Bring your schema. We'll show you what the context layer generates from it on a call.