<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
<title>Eshagh Shahnavazi</title>
<link>https://eshagh.dev</link>
<description>Eshagh — systems engineer, co-founder at Fenext (Istanbul). Production LLM pipelines, RAG and multi-role B2B platforms with explicit authority and audit trails. Next.js, PostgreSQL, pgvector, LiveKit.</description>
<language>en</language>
<atom:link href="https://eshagh.dev/rss.xml" rel="self" type="application/rss+xml" />
<item><title>The confidence gate: automate the reading, not the deciding</title><link>https://eshagh.dev/blog/confidence-gate</link><guid isPermaLink="true">https://eshagh.dev/blog/confidence-gate</guid><description>A confidence gate is the rule set that decides whether a language model&apos;s extraction is allowed to reach a customer without a human. Automate the reading — classification and extraction — and keep a person on the deciding. Anything below the gate still gets acknowledged, then queued with the reason it was held.</description><pubDate>Sun, 26 Jul 2026 00:00:00 GMT</pubDate><category>AI systems</category></item><item><title>Long-term memory that isn&apos;t a chat transcript</title><link>https://eshagh.dev/blog/long-term-memory</link><guid isPermaLink="true">https://eshagh.dev/blog/long-term-memory</guid><description>Storing conversations is not memory. In a care platform for families raising autistic children, memory is three separate tables — what happened, what is generally true, and what has actually worked — with a scheduled job promoting events into patterns, effectiveness computed as arithmetic rather than model opinion, and every answer citing the family&apos;s own record.</description><pubDate>Sun, 26 Jul 2026 00:00:00 GMT</pubDate><category>AI systems</category></item><item><title>Natural-language queries without opening your database</title><link>https://eshagh.dev/blog/nl-to-sql-guardrails</link><guid isPermaLink="true">https://eshagh.dev/blog/nl-to-sql-guardrails</guid><description>Letting recruiters ask questions in plain Turkish means letting a model write SQL against a multi-tenant database. The safety is not in the prompt — it is in a guardrail layer between generation and execution: an allow-listed schema, a forced LIMIT, rejected mutations, and a tenant filter bound as a parameter the model never sees.</description><pubDate>Sun, 26 Jul 2026 00:00:00 GMT</pubDate><category>AI systems</category></item>
</channel>
</rss>