https://0607501b.wangpeng.pages.dev/en/data/Recent content in Data on Zhuifeng NotesHugoenSat, 08 Aug 2026 00:00:00 +0000https://0607501b.wangpeng.pages.dev/en/data/data-governance/Sat, 08 Aug 2026 00:00:00 +0000https://0607501b.wangpeng.pages.dev/en/data/data-governance/<p>Many bank data-governance programs degenerate into &ldquo;fill in the metadata, assign owners.&rdquo; To get results, governance must be built into the data production pipeline.</p> <h2 id="do-two-things-first">Do two things first</h2> <ul> <li><strong>Metadata &amp; lineage</strong>: where a field comes from, which jobs transform it, where it flows — must be tracked automatically.</li> <li><strong>Quality rules built-in</strong>: non-null, uniqueness, consistent definitions act as ETL/lakehouse <em>gates</em>; bad data never lands.</li> </ul> <h2 id="security--encryption">Security &amp; encryption</h2> <ul> <li><strong>Encryption at rest</strong>: sensitive fields (ID numbers, card numbers) encrypted on disk.</li> <li><strong>Dynamic masking</strong>: query-side masking by role; dev environments never see plaintext.</li> </ul> <blockquote> <p>The goal of governance isn&rsquo;t a pretty report — it&rsquo;s giving downstream systems confidence to use the data.</p>