PNPrithwish Nathinprithwish-nath.hashnode.dev·Jul 30 · 16 min readHow Rotating Residential Proxies Work: Per-Request vs Sticky vs TimedTL;DR: This article explains what a rotating residential proxy is, and how rotation is actually three genuinely different behaviors - and not a single toggle - using real exit IP logs. I wrote a prev00
PNPrithwish Nathinprithwish-nath.hashnode.dev·Jul 29 · 23 min readHow LLMs Get Training Data from the WebTL;DR: AI companies don’t train language models on raw HTML. Learn how AI labs fetch, extract, deduplicate, and filter web crawls into a clean, reproducible corpus. So, You Want To Train a Model… Can 00
PNPrithwish Nathinprithwish-nath.hashnode.dev·Jul 21 · 21 min readWhy Increasing Concurrency Causes HTTP 429 Errors (It’s Not Just Too Many Requests)TL;DR: Concurrency is usually introduced as a performance tip. Parallelism is a good thing, so more workers should mean more requests, and more data. If ten workers are good, why not twenty? If twenty00
PNPrithwish Nathinprithwish-nath.hashnode.dev·Jun 25 · 27 min readReducing MCP Response Sizes for LLM Context LimitsA while back I added Bright Data’s MCP to Claude Code, asked it to fetch a docs page, and watched it crash and burn spectacularly: Error: MCP tool “scrape_as_markdown” response (278649 tokens) exceed00
PNPrithwish Nathinprithwish-nath.hashnode.dev·Jun 19 · 23 min readStop Paying the JSON Tax. Build Faster Data Pipelines in Python with Apache Arrow.If you’ve ever built data pipelines for analytics, feature extraction, or model training, you’ve probably noticed a pattern: scraping or ingestion is rarely your bottleneck. It’s that the pipeline tec00
