Your data, described
Every field, join, and model carries a plain-English note on what it means, so a column is never just a column.
refund_rate:
refunded ÷ orders, no exchangesAI on your data
AI can answer anything. The answers you run a business on have to be right.
Most AI tools answer from whatever you paste in. Operator answers from your data joined across tools, every field described in plain English, your business context attached.
Ask in plain English. Get answers, not guesses.
Operator reads from your data joined across tools: your numbers, defined once, used everywhere. When it says margin slipped, it's reading the same definition of margin your dashboards use.
Which subscription plans drove the dip?
Ran Execute SQL · 3 models
select plan, repeat_rate_60d
from subscription_cohort_facts
where month = '2026-05' -- 312 rows · 0.4sThe 2-bag monthly plan. Its 60-day reorder rate dropped 14 points. The 1-bag and wholesale lines held steady.
See where every number came from.
Every answer can show its work: which fields it used, what they mean, which tools they come from, and when that data last synced. Nothing asks to be taken on faith. The full map of how your models connect lives in the modeling layer.
lineage · repeat_rate
$ trace repeat_rate
shopify.orders → orders_base → repeat_rate
sources: shopify ✓ subscriptions ✓ ads ✓
last_sync: 14 minutes ago
defined_in: models/orders.yml · reviewedIt proposes. You approve.
When Operator wants to change how your data is organized, a new field or a remapped rule, the change waits for a yes. Nothing moves under your feet.
models/customers.yml awaiting review
proposed by Operator · just now
acquisition_cohort:
description: first-order month + channel
+ ltv_9_month:
+ description: gross profit per customer, first 9 months
+ type: currencyOperator doesn't just see your data. It understands your world.
Every answer draws on layers of context most AI tools never have. It's all written down, and you can read every line of it.
your-company / context what operator reads
Your business
What you sell, how you make money, which promo ran in March, what changed and when. The facts behind the numbers.
context.promos:
March: 20% off first orderYour industry
Subscription DTC reads differently than B2B services. Operator reasons inside the world you operate in, not a generic one.
context.vertical:
subscription DTCWhat it's learned
Your definitions, your preferences, your weekly routines. It gets sharper about your business, only yours, every time you use it.
memory.brand_spend:
set by campaign name prefixAnd it compounds. Every conversation leaves something behind: a definition pinned down, a caveat worth remembering, a routine learned. The hundredth answer is sharper than the first.
Other AI tools start with a blank brain and hand you the controls.
Switchboard's context arrives full. During setup, data engineers connect your tools and write down what every number means. The platform keeps it current. Operator deepens it with every conversation. By the time you ask your first question, the homework is done.
Book 15-min demoFAQ
How do I know an answer is right?
Ask it to show its work. Every answer can list the fields it used, what they mean, and when the underlying data last synced. If something looks off, you trace it to the source instead of taking it on faith.
Is our data used to train AI models?
No. Operator runs on enterprise AI APIs that don't train on your data — and what it learns about your business stays in your workspace, never shared across customers.
What happens when Operator isn't sure?
It asks. When a question is ambiguous — which margin, which date range — Operator asks before it answers. And bigger moves, like changing how your data is organized, always wait for your approval.
What does "described data" mean?
Behind every chart is a definition: what counts as a repeat order, how refunds are treated, where a "day" starts. Switchboard writes those down in plain English, and Operator reads them before it answers. That's why two people asking the same question get the same number.
Get started
See your own data answer questions.
Book a demo. We'll show you Operator grounded in data shaped like yours.