<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>TCP3 blog</title>
    <link>https://tcp3.ai/blog</link>
    <atom:link href="https://tcp3.ai/rss.xml" rel="self" type="application/rss+xml"/>
    <description>Field notes on building with AI — agents, MCP, and product thinking from TCP3.</description>
    <language>en</language>
    <item>
      <title>Your business doesn&apos;t need another chatbot</title>
      <link>https://tcp3.ai/blog/your-business-doesnt-need-another-chatbot</link>
      <guid isPermaLink="true">https://tcp3.ai/blog/your-business-doesnt-need-another-chatbot</guid>
      <pubDate>Sun, 16 Aug 2026 00:00:00 GMT</pubDate>
      <description>A chatbot talks about the work; an agent finishes it — it books the appointment, updates the calendar, sends the follow-up. If your software still hands every task back to a person, you bought a nicer FAQ, not help. TCP3 builds agents that act on your real systems, not bots that describe what someone else has to do.</description>
      <category>AI agents</category>
      <category>Product thinking</category>
    </item>
    <item>
      <title>What a hosted MCP server actually buys you</title>
      <link>https://tcp3.ai/blog/what-a-hosted-mcp-server-actually-buys-you</link>
      <guid isPermaLink="true">https://tcp3.ai/blog/what-a-hosted-mcp-server-actually-buys-you</guid>
      <pubDate>Sun, 16 Aug 2026 00:00:00 GMT</pubDate>
      <description>MCP is the standard that lets an AI assistant use your app — read data, take actions — instead of just talking about it. Building one for a demo takes an afternoon; running one in production means auth, multi-tenancy, and uptime. A hosted MCP server hands you the connection and keeps that plumbing someone else&apos;s problem.</description>
      <category>MCP</category>
      <category>AI agents</category>
    </item>
  </channel>
</rss>
