A new way to read your business

Ask the question.
Trust the answer.

Pillar's Ask Your Data turns plain-English questions into governed answers — grounded in your curated metric registry, audited end-to-end, and read-only by construction. No mystery numbers in the board deck.

Target Integrated Data Lake Postgres
Claude Sonnet 4.6
Produce a Q3 earnings report for the executive committee

Q3 2026 net income reached $412K (+18% YoY) on $2.84M revenue, with EBITDA up 44% on operating leverage.

Answered Governed 6 rows 4.1 s Cached
Line itemQ3 2026Q3 2025YoY
Revenue$2.84M$2.32M+22.4%
COGS($412K)($358K)+15.1%
Gross Profit$2.43M$1.96M+24.0%
Operating Expenses($1.78M)($1.51M)+17.9%
EBITDA$645K$448K+44.0%
Net Income$412K$349K+18.0%
metric: Revenuemetric: Gross Profitmetric: EBITDAmetric: Net Incomedimension: fiscal_quartertable: fact_pnl
Try next Bridge EBITDA from Q2 to Q3
Ask about revenue, pipeline, support, headcount, cash, or other governed metrics…

Two ways to read the same truth.

Every dashboard pack and every conversational answer pulls from the same curated metric registry. Click through a dashboard for the visual narrative; ask the follow-up in plain English when a number raises a question.

Dashboard pack LIVE

Revenue Pulse

MRR, NRR, churn, and segment performance — refreshed every transformation run.

MRR
$184K
+12%
NRR
118%
+4%
Churn
2.1%
+0.3
Follow-up question Same metric registry
Why is MRR up 12% this month?

MRR grew $19.4K this month, driven primarily by Enterprise segment expansions (+$14.2K) and two new SMB contracts (+$5.2K).

metric: MRR dimension: revenue_segment table: fact_revenue

Same metric. Same source. Asked in 5 seconds, answered in about four seconds.

From question to governed answer in seconds.

Four steps, fully transparent. You see the metric matched, the SQL generated, the latency, and the citations — every time.

You ask

In plain English. No SQL, no ticket queue.

Pillar searches the registry

Looks up the matching governed metric — definition, sources, filters.

Generates read-only SQL

Bound to the marts schema, statement timeout, no writes.

Returns the answer

With citations, latency, row count, and a follow-up suggestion.

Why this isn't another text-to-SQL gimmick.

Most "AI on your data" tools hand a language model raw schema access and hope for the best. Ask Your Data is built on the curated metric registry behind every Pillar dashboard — so the same numbers your board sees are the ones you can ask follow-ups against.

Governed, not guessed

Every answer is grounded in a curated metric from the registry — not a one-shot text-to-SQL guess. The answer card cites the metric, table, and columns used so you can verify in one glance.

Audit-ready by default

Every question, generated SQL, latency, citation, and feedback is persisted. Nothing happens off the record — perfect for finance, compliance, and post-hoc review.

Read-only by construction

Queries run inside a read-only transaction, locked to the marts schema, with statement timeouts and semicolon rejection. No way to mutate data, no surprise side effects.

Ask the question. Trust the answer.

Bring every system into one trusted foundation, ship dashboards that move the business, and let your team ask the follow-ups in plain English.