Bbingkun zhaoOct 24 · 3 min readA Small Data QA Pipeline for 156 Game Walkthrough PagesA library of 156 walkthrough pages can fail in quiet ways. One wrong video ID, a duplicated level, or a guide mapped to a similarly named game may stay online for months because every individual page game-guide-lab.hashnode.dev
RRobin SishodiaOct 23 · 2 min readSalesforce Duplicate & Matching Rules Cheat Sheet: Block, Alert, Merge and Admin Interview Q&A (Day 18/30)Stop messy duplicates before they land, and clean up the ones already in your data. Duplicate and matching rules are core data quality topics in any admin interview. The 4 Building Blocks Block Whsishodiarobin.hashnode.dev
OOsitrusteesOct 23 · 9 min readOsitrustees Engineering Notes: Designing a Reproducible Market Data Research RecordHow structured metadata, source provenance and explicit research conditions help make market-data observations easier to review. A market research result can be difficult to reproduce even when the or1ositrustees.hashnode.dev
RReavidOct 21 · 5 min readWhich rows of a converted bank statement can you trust?Most bookkeepers know the email: "Here are the statements," with a stack of PDFs attached. Sometimes the months are older than the bank's online export window, the account is already closed, or the clstatementxl.hashnode.dev
RRohan MehtaOct 13 · 4 min read6 Ways to Make a Field Required in Salesforce (and When to Use Each)Ask any Salesforce admin what ruins reports fastest, and the answer is almost always the same: incomplete data. Blank fields creep into records, dashboards skew, and automation fires on missing valuesway2force.hashnode.dev
AAnakin TechOct 13 · 8 min readScraped Data Quality Is a Runtime Contract, Not a Cleanup TaskBad scraped data rarely announces itself as bad. It lands in the warehouse looking ordinary, passes through dashboards, and only becomes visible when someone makes a decision from a chart that should 1Oanakinwriters.hashnode.dev
JJorgeOct 10 · 4 min readFabric's Silver Layer: Making Your Data ReliableA successful pipeline does not always mean good data. A pipeline can run without errors and still deliver data you should not trust. That is why the Silver layer deserves more attention than simply “cbuildingagain.hashnode.dev
AAllyn Sharon Laura EugeneOct 5 · 3 min readCan Better Data Lead to Better AI?Why I Became Interested in Data Quality As I became more interested in data science, I started exploring the different ways we can improve machine learning models. However, as I learned about model pedatasciencefordataaiandbeyond.hashnode.dev
SSTAE HIMSep 24 · 6 min readCollecting Google Maps Leads Is Easy. Qualifying Them Is the Real Workflow.Collecting local business records is only the first stage of a useful lead workflow. Before outreach, each record still needs four independent checks: business fit, contactability, an observable oppor11Sstae-him.hashnode.dev
RRakesh RanderiaSep 6 · 4 min readFrom Data Ownership to AI ReadinessAI governance is often discussed as if it starts with the model. In practice, it usually starts much earlier — with the data. If an organisation cannot clearly explain who owns its data, how it is clarakeshranderia.hashnode.dev