<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"><channel>
  <title>Tanveer Riaz - field notes</title>
  <link>https://tanveerriaz.me/blog</link>
  <description>Notes on AI workflows, local-first systems, regulated delivery, and useful products.</description>
  <item>
    <title>SideSpark started with a conversation</title>
    <link>https://tanveerriaz.me/blog/sidespark-started-with-a-conversation</link>
    <guid>https://tanveerriaz.me/blog/sidespark-started-with-a-conversation</guid>
    <description>At the ChatGPT Sites Hackathon in Singapore, Benny and I talked about workplace connections. That conversation became SideSpark: small shared breaks for meeting coworkers and exchanging skills.</description>
  </item>
  <item>
    <title>Mentorship works both ways</title>
    <link>https://tanveerriaz.me/blog/mentorship-works-both-ways</link>
    <guid>https://tanveerriaz.me/blog/mentorship-works-both-ways</guid>
    <description>Four months in Standard Chartered Singapore × Nanyang Technological University Singapore’s LevelUP! programme reminded me that mentoring is an exchange, not a handover.</description>
    <pubDate>Sun, 30 Aug 2026 00:00:00 GMT</pubDate>
  </item>
  <item>
    <title>Two lights and a thread</title>
    <link>https://tanveerriaz.me/blog/87k-windows</link>
    <guid>https://tanveerriaz.me/blog/87k-windows</guid>
    <description>Building 87K Windows at the 65 Labs Gemini hackathon: local Gemma as a privacy layer, a deliberately boring matcher, and a Gemini that opens the door then gets out of the way.</description>
    <pubDate>Sun, 23 Aug 2026 00:00:00 GMT</pubDate>
  </item>
  <item>
    <title>Agents don&apos;t clock out</title>
    <link>https://tanveerriaz.me/blog/agents-dont-clock-out</link>
    <guid>https://tanveerriaz.me/blog/agents-dont-clock-out</guid>
    <description>The Claude community met in Singapore. Instead of a photo dump, the recap became a 15-second 8-bit story generated with fal.ai from real event photos - and then a reusable skill.</description>
    <pubDate>Sat, 08 Aug 2026 00:00:00 GMT</pubDate>
  </item>
  <item>
    <title>Codex gave developers pets. I gave the pets a leaderboard.</title>
    <link>https://tanveerriaz.me/blog/dashboard-atlas</link>
    <guid>https://tanveerriaz.me/blog/dashboard-atlas</guid>
    <description>The build story behind Codex Pet League: one wrong first image, authentic community pets, a live Vercel dashboard, four Chroa design experiments, and every prompt used.</description>
    <pubDate>Sun, 02 Aug 2026 00:00:00 GMT</pubDate>
  </item>
  <item>
    <title>Three doors into one local agent</title>
    <link>https://tanveerriaz.me/blog/hermes-gbrain</link>
    <guid>https://tanveerriaz.me/blog/hermes-gbrain</guid>
    <description>How terminal, Telegram, and WhatsApp sessions use one deliberately bounded local memory layer—with model routing, privacy boundaries, and a public-safe architecture.</description>
    <pubDate>Sun, 02 Aug 2026 00:00:00 GMT</pubDate>
  </item>
  <item>
    <title>A graph is not a guide: building Ishara</title>
    <link>https://tanveerriaz.me/blog/ishara</link>
    <guid>https://tanveerriaz.me/blog/ishara</guid>
    <description>How Ishara became a meaning-first Qur’an study path through explicit node identity, human labels, contextual verses, and lighter first-click payloads.</description>
    <pubDate>Sun, 02 Aug 2026 00:00:00 GMT</pubDate>
  </item>
  <item>
    <title>What 20 years in banking taught me about shipping AI</title>
    <link>https://tanveerriaz.me/blog/banking-shipping-ai</link>
    <guid>https://tanveerriaz.me/blog/banking-shipping-ai</guid>
    <description>The bridge between global banking delivery discipline and after-hours AI building - regulated instincts, sprint rigor, and why the harness matters more than the model.</description>
    <pubDate>Sat, 04 Jul 2026 00:00:00 GMT</pubDate>
  </item>
  <item>
    <title>You don&apos;t need a $240/year subscription to use AI</title>
    <link>https://tanveerriaz.me/blog/local-ai</link>
    <guid>https://tanveerriaz.me/blog/local-ai</guid>
    <description>Field notes from a live session on running capable AI on your own machine - private, offline, and free. Ollama, AnythingLLM, local agents, and a real business automation flow.</description>
    <pubDate>Sat, 04 Jul 2026 00:00:00 GMT</pubDate>
  </item>
  <item>
    <title>Same flight, two endings</title>
    <link>https://tanveerriaz.me/blog/same-flight-two-endings</link>
    <guid>https://tanveerriaz.me/blog/same-flight-two-endings</guid>
    <description>Tracking an overnight family flight two ways - a polished native AI scheduler vs. my own agent over Telegram.</description>
    <pubDate>Sat, 27 Jun 2026 00:00:00 GMT</pubDate>
  </item>
  <item>
    <title>The harness matters</title>
    <link>https://tanveerriaz.me/blog/harness-matters</link>
    <guid>https://tanveerriaz.me/blog/harness-matters</guid>
    <description>The model isn</description>
    <pubDate>Sat, 20 Jun 2026 00:00:00 GMT</pubDate>
  </item>
  <item>
    <title>&quot;Which AI model is the best?&quot;</title>
    <link>https://tanveerriaz.me/blog/which-ai-model</link>
    <guid>https://tanveerriaz.me/blog/which-ai-model</guid>
    <description>It</description>
    <pubDate>Sun, 31 May 2026 00:00:00 GMT</pubDate>
  </item>
</channel></rss>
