SSpinHireinspinhire.hashnode.dev·4h ago · 6 min readMontana is not Malta: what we learned normalizing thousands of job postingsWe run a niche job index: once a day a worker pulls vacancies from company career pages and applicant tracking systems, cleans them up and publishes them. It sounds like a solved problem. In practice,00
SKShubham Kumar Singhinblog.realdev.club·1d ago · 17 min readHow We Built a Data Pipeline: From Raw 2.8M Messy Tweets to Training DataEveryone wants to talk about the model. Nobody wants to talk about the CSV. I set out to build an AI customer support agent — the kind that reads a message, figures out what's wrong, and drafts a grou00
ATAnakin Techinanakinwriters.hashnode.dev·1d ago · 7 min readFood Delivery Web Data Is an Operations System, Not a DashboardMost food delivery data programs fail because they are built like reporting projects. A dashboard that shows yesterday's competitor prices is not intelligence, it is a historical artifact. In a market01O
CDCoding Dropletsincodingdroplets.com·2d ago · 13 min readHow to Build Incremental RAG Indexing in .NET: A Real-World WalkthroughIncremental RAG indexing is what separates a demo knowledge base from one you can still trust six months later. The first ingestion run is easy: read every document, chunk it, embed it, write it to a 00
NNumberCheckerinnumberchecker.hashnode.dev·2d ago · 5 min readHandling ambiguity in multi-format numerical identifier ingestionIn high-throughput ingestion systems, the most common point of failure is not the database or the network, but the "garbage in" problem at the ingestion layer. When building a service that consumes nu00
CCheckNumberinchecknumberofficial.hashnode.dev·2d ago · 5 min readRefactoring Legacy Financial Document Parsers for Modern Neural Extraction ArchitecturesThe Architectural Shift in Financial Document Processing Financial document processing systems often begin as collections of rigid, coordinate-based parsers. When dealing with structured documents lik00
DADamilare Agbainagbad.hashnode.dev·3d ago · 6 min readThree bugs that passed every test I hadA market snapshot split across two HTTP calls, a pandas timestamp cast that assumed the wrong resolution, and an FX session boundary I was certain about — three bugs a green test suite was perfectly h00
CCheckNumberinchecknumberofficial.hashnode.dev·3d ago · 5 min readMigrating Legacy Rule-Based Financial Document Parsers to Neural Extraction ArchitecturesThe Architectural Shift: Moving from Regex to Neural Extraction For many financial engineering teams, the "legacy" stack for document processing is built on a foundation of coordinate-based mapping an00
SCSriharsha Chintalapaniinblog.getcollate.io·3d ago · 18 min readWhy We Built a Native Agentic Harness for Data on Top of Collate's Context PlatformClaude Code and Codex are usually described as coding agents. That description is accurate, but it leaves out a large part of why they are useful. A strong model is clearly important, but the model is00
NNumberCheckerinnumberchecker.hashnode.dev·4d ago · 5 min readThe complexity of state-based validation for non-standard numerical identifiers — 20260914The Complexity of State-Based Validation for Non-Standard Numerical Identifiers In high-throughput data ingestion pipelines, the initial stage of processing often involves sanitizing and validating in00
