DATA OPS
Hybrid regex + LLM PII classifier for Snowflake columns
Runs cheap regex screening over sampled Snowflake column values and escalates ambiguous hits to an LLM for category and confidence.
How it runs
The automated pipeline, trigger to output.
- TriggerScheduled classifier run
- ActionSample candidate Snowflake columnsSnowflake
- LogicRegex screen; isolate ambiguous samples
- ActionLLM adjudicates gray-area samplesOpenAI
- LogicGrade by confidence: high/medium/low
- ActionQuarantine high-confidence tablesSnowflake
- OutputFile graded Linear ticketLinear
What it does
Samples values from new or recently changed Snowflake columns and screens them with fast regex rules for structured PII like card numbers, SSNs, and emails. Columns that partially match or look like free-text names get escalated to an LLM that returns a PII category, confidence score, and one-line rationale. High-confidence findings quarantine the table; medium ones open a Linear ticket without locking; low ones are dropped.
When to use it
Use it when pure regex throws too many false positives on messy columns (free-text notes, mixed addresses) and you want an LLM to adjudicate only the gray-area cases without paying to classify every value.
How it works
- 1A schedule starts the scan.
- 2Sample values from candidate Snowflake columns.
- 3Run regex screening; route clean hits and misses directly.
- 4Send only ambiguous samples to the LLM for category and confidence.
- 5Branch on the combined confidence: high quarantines, medium tickets, low drops.
- 6Revoke SELECT on high-confidence tables and file a graded Linear ticket.
Set it up
What you configure once, before turning it on.
- 1Connect SnowflakeWarehouses, queries, shares.
- 2Connect OpenAIModels, embeddings, files.
- 3Connect LinearIssues, projects, cycles, triage.
- 4Set each agent's modelWe leave models unset so you pick the tier — fast + cheap, or top-quality.
- 5Tune it to your dataEdit the prompts, filters, and field mappings so it matches how your team works.
- 6Test, then turn it onRun once against a sample, confirm the output, then enable the trigger.
More Data Ops workflows
BigQuery Per-Team Budget Breach Alert to PagerDuty
Tracks month-to-date BigQuery scheduled-query spend per team and, when a team crosses its monthly budget, pages the team's on-call in PagerDuty and snapshots the spend breakdown…
dbt orphan model detector with Linear cleanup tickets
Scans your dbt manifest for models that no other model, exposure, or BI tool consumes.
Weekly BigQuery Cost Trend Sheet and Exec Digest
Compiles week-over-week BigQuery scheduled-query cost by owner and dataset into a Google Sheet with trend columns.
Backfill Missing Owner Labels on BigQuery Scheduled Queries
Finds scheduled queries with no owner label, infers the likely owner from creator metadata and target-table lineage, proposes a label.
Daily BigQuery Scheduled-Query Cost Attribution to Owners
Each morning, totals the prior day's on-demand bytes-billed per scheduled query, maps each query to its owner from a label, and posts a per-owner cost leaderboard to Slack.
dbt source freshness watcher with severity-routed alerts
Checks Snowflake loaded-at timestamps against each dbt source's freshness SLA, then routes warnings to Slack and hard breaches to a PagerDuty incident so stale data never…
Run it inside a business
This workflow drops into a full company template. Import the org, and this is one of the playbooks its agents run.
Run this workflow in your colony.
14-day trial. No DevOps. No Sales call. Provisioned in under a minute.