<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Architecture on Solvinc — AI's Uncharted Edge</title><link>https://blog.solvinc.io/tags/architecture/</link><description>Recent content in Architecture on Solvinc — AI's Uncharted Edge</description><generator>Hugo -- 0.139.0</generator><language>en-us</language><lastBuildDate>Tue, 17 Nov 2026 00:00:00 +0700</lastBuildDate><atom:link href="https://blog.solvinc.io/tags/architecture/index.xml" rel="self" type="application/rss+xml"/><item><title>The Price Was Never in the HTML: Font Obfuscation, and the Cracker That Recovers Every Glyph</title><link>https://blog.solvinc.io/posts/font-obfuscation-glyph-mapping-cracker/</link><pubDate>Tue, 17 Nov 2026 00:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/font-obfuscation-glyph-mapping-cracker/</guid><description>Sites ship randomized fonts so your scraper stores garbage instead of prices. I built the obfuscator, cracked it: 11/11 glyphs recovered, code included.</description></item><item><title>The Durable Queue You Already Own: Exactly-Once Effects From SQLite, Zero Infrastructure</title><link>https://blog.solvinc.io/posts/sqlite-durable-job-queue-lease-idempotency-dead-letter/</link><pubDate>Mon, 16 Nov 2026 09:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/sqlite-durable-job-queue-lease-idempotency-dead-letter/</guid><description>Build a crash-proof job queue on SQLite: leases, idempotent enqueue, retry backoff, dead-letters. Tested: 4 SIGKILLed workers, 0 lost jobs, 5,673 jobs/s.</description></item><item><title>I Replaced a $1,200/yr SERP API With a $6/mo VPS: the Tiered-Engine Architecture, the Ban Math, and Where It Loses</title><link>https://blog.solvinc.io/posts/self-hosted-searxng-serp-api-replacement/</link><pubDate>Sat, 14 Nov 2026 00:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/self-hosted-searxng-serp-api-replacement/</guid><description>Build a $0 SERP API on self-hosted SearXNG: tiered engine pools, ban math, dedup + fusion ranking, and the honest benchmarks where it loses to paid APIs.</description></item><item><title>Your Scraper Doesn't Need More Retries. It Needs a Congestion Controller.</title><link>https://blog.solvinc.io/posts/adaptive-throttling-aimd-circuit-breaker-scrapers/</link><pubDate>Thu, 12 Nov 2026 00:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/adaptive-throttling-aimd-circuit-breaker-scrapers/</guid><description>Scrapers die by 403, not 429. A tested client-side AIMD controller and circuit breaker that converges on the target&amp;#39;s real rate limit instead of guessing.</description></item><item><title>The Hedge Budget: Google's Tail-Fix for Latency, Priced So It Doesn't Melt Your API Bill</title><link>https://blog.solvinc.io/posts/hedge-budget-tail-latency-api-cost-python/</link><pubDate>Tue, 10 Nov 2026 00:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/hedge-budget-tail-latency-api-cost-python/</guid><description>A tested Python hedging controller that cuts tail latency like Tail-at-Scale, with a hard budget gate so duplicate requests never blow your API spend.</description></item><item><title>The $6.85 Question: A Cost Engine That Tells You Whether to Buy a Scraping API, Roll Your Own, or Bolt a Captcha Solver Onto a Weak One</title><link>https://blog.solvinc.io/posts/captcha-antibot-cost-engine-solver-retry-python/</link><pubDate>Sun, 08 Nov 2026 00:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/captcha-antibot-cost-engine-solver-retry-python/</guid><description>Anti-bot is a cost-engineering problem, not a hacking problem. A tested Monte Carlo engine that prices per-successful-row across APIs, solver retries, and risk.</description></item><item><title>The 97% Model Cascade: Route Every LLM Call to the Cheapest Model That Can Actually Do the Job</title><link>https://blog.solvinc.io/posts/llm-cost-cascade-router-python/</link><pubDate>Sun, 08 Nov 2026 00:00:00 +0700</pubDate><guid>https://blog.solvinc.io/posts/llm-cost-cascade-router-python/</guid><description>A tested Python cascade router that cuts LLM spend by routing cheap-first and escalating only on graded failure, with the math on when it backfires.</description></item></channel></rss>