Pilot-ready, not pilot-validated yet

Business Data Pilot Offer

Turn the project from a portfolio demo into a safe pilot-ready offer for people who can bring anonymized business-shaped data or a real data-quality workflow problem.

Step 1

Scope a safe dataset

Dataset shape, non-sensitive field names, business rule candidates, and permission boundary.

Step 2

Run a bounded replay

CSV upload or read-only PostgreSQL run with row/column limits and no stored raw data.

Step 3

Review evidence-backed findings

Useful finding, missed rule, confusing recommendation, or accepted root-cause hypothesis.

Step 4

Publish redacted evidence

GitHub issue with permission to count, no private data, and a concrete observed result.

Eligible data sources

  • anonymized order, ticket, transaction, inventory, or signup CSV
  • read-only PostgreSQL table with non-sensitive columns
  • synthetic-but-business-shaped export that mirrors a real workflow
  • written business case when data cannot be shared

Evidence gates

  • non-owner reviewer or pilot contact
  • public GitHub issue or public review link
  • explicit permission to count the redacted result
  • dataset shape and path tried
  • one concrete useful, confusing, missing, or reproducible finding
  • confirmation that no private data was posted

Current public counts

Submission paths

Published a pilot-ready business data offer with 4 pilot steps, 4 eligible data-source types, 6 evidence gates, a public pilot issue, and zero current external pilot claims.