Generate verified, read-only Postgres queries from plain English

Question: "Who are our top 10 customers by revenue last quarter?"

SELECT
  c.customer_id,
  c.company_name,
  SUM(o.total_amount) AS revenue
FROM customers c
JOIN orders o ON o.customer_id = c.customer_id
WHERE o.order_date >= DATE_TRUNC('quarter', CURRENT_DATE - INTERVAL '3 months')
  AND o.order_date <  DATE_TRUNC('quarter', CURRENT_DATE)
GROUP BY c.customer_id, c.company_name
ORDER BY revenue DESC
LIMIT 10;

Generated by Chion's 13-step pipeline. Read-only SELECT, LIMIT enforced, schema-validated.

Type your question, Chion profiles your schema, builds a SQL contract, generates SQL inside that contract, validates it through L1 (read-only) and L2 (runtime lint), runs it read-only with LIMIT enforced, and renders the chart. Every step is auditable.

Ask "top 10 customers by revenue" and you get TopK ranking, which picks a different SQL shape than "compare revenue by region over time," which routes to comparison. The strategy determines the query structure before the model writes anything.

• entity_lookup
• comparison
• topk_ranked
• extrema_detection
• dimension_breakdown
• universal_quantifier
• time_bounded_only

Read-only SELECT. AES-256-GCM credential vault. 1,000-row cap. Immutable audit log. Read the full security model →

A ranking query, a multi-CTE ratio, and a window-function running total: three of the most common patterns Chion generates.

SELECT
  c.customer_id,
  c.company_name,
  SUM(o.total_amount) AS revenue
FROM customers c
JOIN orders o ON o.customer_id = c.customer_id
WHERE o.order_date >= DATE_TRUNC('quarter', CURRENT_DATE - INTERVAL '3 months')
  AND o.order_date < DATE_TRUNC('quarter', CURRENT_DATE)
GROUP BY c.customer_id, c.company_name
ORDER BY revenue DESC
LIMIT 10;

WITH monthly AS (
  SELECT
    DATE_TRUNC('month', canceled_at) AS month,
    COUNT(*) AS churned
  FROM subscriptions
  WHERE canceled_at IS NOT NULL
  GROUP BY 1
),
active AS (
  SELECT
    DATE_TRUNC('month', period_start) AS month,
    COUNT(DISTINCT user_id) AS active
  FROM subscriptions
  GROUP BY 1
)
SELECT
  m.month,
  m.churned,
  a.active,
  ROUND(m.churned::numeric / NULLIF(a.active, 0) * 100, 1) AS churn_rate_pct
FROM monthly m
JOIN active a ON a.month = m.month
ORDER BY m.month;

WITH weekly AS (
  SELECT DATE_TRUNC('week', order_date) AS week, SUM(amount) AS revenue
  FROM orders GROUP BY 1
)
SELECT
  week,
  revenue,
  SUM(revenue) OVER (ORDER BY week) AS running_total
FROM weekly
ORDER BY week;

A standalone SQL query generator gives you a query: one off, ad-hoc, regenerated next time you ask. Chion promotes each verified query into a per-persona SQL skill library, scoped to the persona that owns the data. Finance personas inherit revenue queries; ops personas inherit logistics queries. The library compounds with every question your team asks.