<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Bharateesha Lvn · Blog</title><description>CS grad · building ML systems &amp; the infra that runs them</description><link>https://bharateesha.com/</link><language>en-us</language><item><title>When a Machine Speaks.</title><link>https://bharateesha.com/blog/when-a-machine-speaks/</link><guid isPermaLink="true">https://bharateesha.com/blog/when-a-machine-speaks/</guid><description>What a production cascaded STT–LLM–TTS architecture actually looks like — from sound hitting a microphone to the reply coming back, including the orchestration layer that turns out to be the hardest part of all.</description><pubDate>Mon, 22 Jun 2026 00:00:00 GMT</pubDate><category>voice-ai</category><category>ml-systems</category><category>architecture</category></item></channel></rss>