Column-Level Lineage Tool — Trace SQL Data Flow Precisely

Table-level lineage tells you “this query reads from table A and writes to table B.” Column-level lineage tells you exactly which columns in B came from which columns in A, and how they were transformed. That precision matters.

Why Column-Level Matters

Impact analysis: “If I drop customer_email from the source table, which downstream reports break?” Table-level lineage can’t answer this. Column-level can.

GDPR compliance: Data privacy regulations require knowing exactly where PII flows. Column-level lineage maps the path of email, phone, and address through your SQL transformations.

Metric debugging: When a KPI looks wrong, trace the calculation back to its source columns. See every JOIN, filter, and aggregation along the way.

FlowScope: Precise SQL Lineage

FlowScope is PondPilot’s SQL lineage engine, built in Rust and compiled to WebAssembly:

  • Column-level granularity: Maps individual column dependencies through CTEs, subqueries, JOINs, and UNIONs
  • Multi-dialect: Supports multiple SQL dialects, not just one vendor’s syntax
  • Client-side: Runs in your browser — your SQL never leaves your machine
  • Open source: Inspect, contribute, or integrate into your own tools

How It Works

Paste a SQL statement into FlowScope. It parses the query, resolves aliases and CTEs, and produces a lineage graph showing how each output column traces back to source columns.

SELECT
  o.order_id,
  c.name as customer_name,
  SUM(oi.quantity * oi.unit_price) as total_value
FROM orders o
JOIN customers c ON o.customer_id = c.id
JOIN order_items oi ON o.order_id = oi.order_id
GROUP BY o.order_id, c.name;

FlowScope maps: total_value ← oi.quantity + oi.unit_price. customer_name ← c.name. Precisely.

Integrate with Your Pipeline

FlowScope’s WASM build means you can embed it in CI/CD pipelines, documentation tools, or internal platforms. Analyze lineage programmatically without running a server.

Try It

Open FlowScope and paste your SQL.

Back to home