luke/ai-podcast
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Clip pipeline improvements, direct YouTube upload, hero redesign, how-it-works updates

Browse Source 
- make_clips: migrate refine_clip_timestamps to mlx-whisper, add LLM caption
  polishing, fix speaker label reversal in grouped caption lines
- upload_clips: interactive episode/clip/platform menus, direct YouTube Shorts
  upload via Data API v3 (bypasses Postiz), direct Bluesky upload
- Website hero: centered layout with left-column cover art on desktop, compact
  text links instead of pill buttons, scaled up typography
- How-it-works: move anatomy section above diagram, update stats (320 names,
  189+ personality layers, 20 towns, 570+ topics, 1400+ scenarios), add
  drunk/high/unhinged callers, voicemails, MLX Whisper GPU, LLM-polished captions
- All footers: add System Status link, remove Ko-fi branding
- .gitignore: YouTube OAuth credential files

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
tcpsyn
2026-02-16 04:06:23 -07:00
parent 3164a70e48
commit f0271e61df
9 changed files with 591 additions and 266 deletions
Download Patch File Download Diff File Expand all files Collapse all files

225
website/how-it-works.html
View File

@@ -69,6 +69,109 @@
<p class="page-subtitle">Every caller on the show is a one-of-a-kind character — generated in real time by a custom-built AI system. Here's a peek behind the curtain.</p>
</section>
<!-- Steps -->
<section class="hiw-section">
<h2>The Anatomy of an AI Caller</h2>
<div class="hiw-steps">
<div class="hiw-step">
<div class="hiw-step-number">1</div>
<div class="hiw-step-content">
<h3>A Person Is Born</h3>
<p>Every caller starts as a blank slate. The system generates a complete identity: name, age, job, hometown, and personality. Each caller gets a unique speaking style — some ramble, some are blunt, some deflect with humor. They have relationships, vehicles, strong food opinions, nostalgic memories, and reasons for being up this late. They know what they were watching on TV, what errand they ran today, and what song was on the radio before they called.</p>
<p>Some callers become regulars. The system tracks returning callers across episodes — they remember past conversations, reference things they talked about before, and their stories evolve over time. You'll hear Carla update you on her divorce, or Carl check in about his gambling recovery. They're not reset between shows.</p>
<p>And some callers are drunk, high, or flat-out unhinged. They'll call with conspiracy theories about pigeons being government drones, existential crises about whether fish know they're wet, or to confess they accidentally set their kitchen on fire trying to make grilled cheese at 3 AM.</p>
<div class="hiw-detail-grid">
<div class="hiw-detail">
<span class="hiw-detail-label">Unique Names</span>
<span class="hiw-detail-value">320</span>
</div>
<div class="hiw-detail">
<span class="hiw-detail-label">Personality Layers</span>
<span class="hiw-detail-value">189+</span>
</div>
<div class="hiw-detail">
<span class="hiw-detail-label">Towns with Real Knowledge</span>
<span class="hiw-detail-value">20</span>
</div>
<div class="hiw-detail">
<span class="hiw-detail-label">Returning Regulars</span>
<span class="hiw-detail-value">12 callers</span>
</div>
</div>
</div>
</div>
<div class="hiw-step">
<div class="hiw-step-number">2</div>
<div class="hiw-step-content">
<h3>They Know Their World</h3>
<p>Callers know real facts about where they live — the restaurants, the highways, the local gossip. When a caller says they're from Lordsburg, they actually know about the Shakespeare ghost town and the drive to Deming. They know the current weather outside their window, what day of the week it is, whether it's monsoon season or chile harvest. They have strong opinions about where to get the best green chile and get nostalgic about how their town used to be. The system also pulls in real-time news so callers can reference things that actually happened today.</p>
</div>
</div>
<div class="hiw-step">
<div class="hiw-step-number">3</div>
<div class="hiw-step-content">
<h3>They Have a Reason to Call</h3>
<p>Some callers have a problem — a fight with a neighbor, a situation at work, something weighing on them at 2 AM. Others call to geek out about Severance, argue about poker strategy, or share something they read about quantum physics. The system draws from over 570 discussion topics across dozens of categories and more than 1,400 life scenarios. Every caller has a purpose, not just a script.</p>
<div class="hiw-split-stat">
<div class="hiw-stat">
<span class="hiw-stat-number">70%</span>
<span class="hiw-stat-label">Need advice</span>
</div>
<div class="hiw-stat">
<span class="hiw-stat-number">30%</span>
<span class="hiw-stat-label">Want to talk about something</span>
</div>
</div>
</div>
</div>
<div class="hiw-step">
<div class="hiw-step-number">4</div>
<div class="hiw-step-content">
<h3>The Conversation Is Real</h3>
<p>Luke talks to each caller using push-to-talk, just like a real radio show. His voice is transcribed in real time, sent to an AI that responds in character, and then converted to speech using a voice engine — all in a few seconds. The AI doesn't just answer questions; it reacts, gets emotional, goes on tangents, and remembers what was said earlier in the show. Callers even react to previous callers — "Hey Luke, I heard that guy Tony earlier and I got to say, he's full of it." It makes the show feel like a living community, not isolated calls.</p>
</div>
</div>
<div class="hiw-step">
<div class="hiw-step-number">5</div>
<div class="hiw-step-content">
<h3>Real Callers Call In Too</h3>
<p>When you dial 208-439-LUKE, your call goes into a live queue. Luke sees you waiting and can take your call right from the control room. Your voice streams in real time — no pre-recording, no delay. You're live on the show, talking to Luke, and the AI callers might even react to what you said. And if Luke isn't live, you can leave a voicemail — it gets transcribed and may get played on a future episode.</p>
</div>
</div>
<div class="hiw-step">
<div class="hiw-step-number">6</div>
<div class="hiw-step-content">
<h3>The Control Room</h3>
<p>The entire show runs through a custom-built control panel. Luke manages callers, plays music and sound effects, runs ads, monitors the call queue, and controls everything from one screen. Audio is routed across multiple channels simultaneously — caller voices, music, sound effects, and live phone audio all on separate tracks. The website shows a live on-air indicator so listeners know when to call in.</p>
<div class="hiw-detail-grid">
<div class="hiw-detail">
<span class="hiw-detail-label">Audio Channels</span>
<span class="hiw-detail-value">5 independent</span>
</div>
<div class="hiw-detail">
<span class="hiw-detail-label">Caller Slots</span>
<span class="hiw-detail-value">10 per session</span>
</div>
<div class="hiw-detail">
<span class="hiw-detail-label">Phone System</span>
<span class="hiw-detail-value">VoIP + WebSocket</span>
</div>
<div class="hiw-detail">
<span class="hiw-detail-label">Live Status</span>
<span class="hiw-detail-value">Real-time CDN</span>
</div>
</div>
</div>
</div>
</div>
</section>
<!-- Overview -->
<section class="hiw-section">
<div class="hiw-card hiw-hero-card">
@@ -211,6 +314,12 @@
</div>
<span>Social Clips</span>
</div>
<div class="diagram-box">
<div class="diagram-icon">
<svg viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2"><path d="M22 12h-4l-3 9L9 3l-3 9H2"/></svg>
</div>
<span>Monitoring</span>
</div>
</div>
<div class="diagram-arrow">&#8595;</div>
<!-- Row 7: Distribution -->
@@ -281,108 +390,6 @@
</div>
</section>
<!-- Steps -->
<section class="hiw-section">
<h2>The Anatomy of an AI Caller</h2>
<div class="hiw-steps">
<div class="hiw-step">
<div class="hiw-step-number">1</div>
<div class="hiw-step-content">
<h3>A Person Is Born</h3>
<p>Every caller starts as a blank slate. The system generates a complete identity: name, age, job, hometown, and personality. Each caller gets a unique speaking style — some ramble, some are blunt, some deflect with humor. They have relationships, vehicles, strong food opinions, nostalgic memories, and reasons for being up this late. They know what they were watching on TV, what errand they ran today, and what song was on the radio before they called.</p>
<p>Some callers become regulars. The system tracks returning callers across episodes — they remember past conversations, reference things they talked about before, and their stories evolve over time. You'll hear Carla update you on her divorce, or Carl check in about his gambling recovery. They're not reset between shows.</p>
<div class="hiw-detail-grid">
<div class="hiw-detail">
<span class="hiw-detail-label">Unique Names</span>
<span class="hiw-detail-value">160 names</span>
</div>
<div class="hiw-detail">
<span class="hiw-detail-label">Personality Layers</span>
<span class="hiw-detail-value">30+</span>
</div>
<div class="hiw-detail">
<span class="hiw-detail-label">Towns with Real Knowledge</span>
<span class="hiw-detail-value">32</span>
</div>
<div class="hiw-detail">
<span class="hiw-detail-label">Returning Regulars</span>
<span class="hiw-detail-value">12+ callers</span>
</div>
</div>
</div>
</div>
<div class="hiw-step">
<div class="hiw-step-number">2</div>
<div class="hiw-step-content">
<h3>They Know Their World</h3>
<p>Callers know real facts about where they live — the restaurants, the highways, the local gossip. When a caller says they're from Lordsburg, they actually know about the Shakespeare ghost town and the drive to Deming. They know the current weather outside their window, what day of the week it is, whether it's monsoon season or chile harvest. They have strong opinions about where to get the best green chile and get nostalgic about how their town used to be. The system also pulls in real-time news so callers can reference things that actually happened today.</p>
</div>
</div>
<div class="hiw-step">
<div class="hiw-step-number">3</div>
<div class="hiw-step-content">
<h3>They Have a Reason to Call</h3>
<p>Some callers have a problem — a fight with a neighbor, a situation at work, something weighing on them at 2 AM. Others call to geek out about Severance, argue about poker strategy, or share something they read about quantum physics. Every caller has a purpose, not just a script.</p>
<div class="hiw-split-stat">
<div class="hiw-stat">
<span class="hiw-stat-number">70%</span>
<span class="hiw-stat-label">Need advice</span>
</div>
<div class="hiw-stat">
<span class="hiw-stat-number">30%</span>
<span class="hiw-stat-label">Want to talk about something</span>
</div>
</div>
</div>
</div>
<div class="hiw-step">
<div class="hiw-step-number">4</div>
<div class="hiw-step-content">
<h3>The Conversation Is Real</h3>
<p>Luke talks to each caller using push-to-talk, just like a real radio show. His voice is transcribed in real time, sent to an AI that responds in character, and then converted to speech using a voice engine — all in a few seconds. The AI doesn't just answer questions; it reacts, gets emotional, goes on tangents, and remembers what was said earlier in the show. Callers even react to previous callers — "Hey Luke, I heard that guy Tony earlier and I got to say, he's full of it." It makes the show feel like a living community, not isolated calls.</p>
</div>
</div>
<div class="hiw-step">
<div class="hiw-step-number">5</div>
<div class="hiw-step-content">
<h3>Real Callers Call In Too</h3>
<p>When you dial 208-439-LUKE, your call goes into a live queue. Luke sees you waiting and can take your call right from the control room. Your voice streams in real time — no pre-recording, no delay. You're live on the show, talking to Luke, and the AI callers might even react to what you said.</p>
</div>
</div>
<div class="hiw-step">
<div class="hiw-step-number">6</div>
<div class="hiw-step-content">
<h3>The Control Room</h3>
<p>The entire show runs through a custom-built control panel. Luke manages callers, plays music and sound effects, runs ads, monitors the call queue, and controls everything from one screen. Audio is routed across multiple channels simultaneously — caller voices, music, sound effects, and live phone audio all on separate tracks. The website shows a live on-air indicator so listeners know when to call in.</p>
<div class="hiw-detail-grid">
<div class="hiw-detail">
<span class="hiw-detail-label">Audio Channels</span>
<span class="hiw-detail-value">5 independent</span>
</div>
<div class="hiw-detail">
<span class="hiw-detail-label">Caller Slots</span>
<span class="hiw-detail-value">10 per session</span>
</div>
<div class="hiw-detail">
<span class="hiw-detail-label">Phone System</span>
<span class="hiw-detail-value">VoIP + WebSocket</span>
</div>
<div class="hiw-detail">
<span class="hiw-detail-label">Live Status</span>
<span class="hiw-detail-value">Real-time CDN</span>
</div>
</div>
</div>
</div>
</div>
</section>
<!-- Post-Production Pipeline -->
<section class="hiw-section">
<h2>From Live Show to Podcast</h2>
@@ -444,11 +451,11 @@
<div class="hiw-step-number">9</div>
<div class="hiw-step-content">
<h3>Automated Publishing</h3>
<p>A single command takes a finished episode and handles everything: the audio is transcribed using speech recognition to generate full-text transcripts, then an LLM analyzes the transcript to write the episode title, description, and chapter markers with timestamps. The episode is uploaded to the podcast server, chapters and transcripts are attached to the metadata, and all media is synced to a global CDN so listeners everywhere get fast downloads.</p>
<p>A single command takes a finished episode and handles everything: the audio is transcribed using MLX Whisper running on Apple Silicon GPU to generate full-text transcripts, then an LLM analyzes the transcript to write the episode title, description, and chapter markers with timestamps. The episode is uploaded to the podcast server, chapters and transcripts are attached to the metadata, and all media is synced to a global CDN so listeners everywhere get fast downloads.</p>
<div class="hiw-detail-grid">
<div class="hiw-detail">
<span class="hiw-detail-label">Transcription</span>
<span class="hiw-detail-value">Whisper AI</span>
<span class="hiw-detail-value">MLX Whisper (GPU)</span>
</div>
<div class="hiw-detail">
<span class="hiw-detail-label">Metadata</span>
@@ -470,7 +477,7 @@
<div class="hiw-step-number">10</div>
<div class="hiw-step-content">
<h3>Automated Social Clips</h3>
<p>No manual editing, no scheduling tools. After each episode, an LLM reads the full transcript and picks the best moments — funny exchanges, wild confessions, heated debates. Each clip is automatically extracted, captioned with word-level timing, and rendered as a vertical video with the show's branding. A second LLM pass writes platform-specific descriptions and hashtags. Then a single script blasts every clip to Instagram Reels, YouTube Shorts, Facebook, Bluesky, and Mastodon simultaneously — six platforms, zero manual work.</p>
<p>No manual editing, no scheduling tools. After each episode, an LLM reads the full transcript and picks the best moments — funny exchanges, wild confessions, heated debates. Each clip is automatically extracted, transcribed with word-level timestamps, then polished by a second LLM pass that fixes punctuation, capitalization, and misheard words while preserving timing. The clips are rendered as vertical video with speaker-labeled captions and the show's branding. A third LLM writes platform-specific descriptions and hashtags. Then clips are uploaded directly to YouTube Shorts and Bluesky via their APIs, and pushed to Instagram Reels, Facebook, and Mastodon — six platforms, zero manual work.</p>
<div class="hiw-detail-grid">
<div class="hiw-detail">
<span class="hiw-detail-label">Human Effort</span>
@@ -482,7 +489,7 @@
</div>
<div class="hiw-detail">
<span class="hiw-detail-label">Captions</span>
<span class="hiw-detail-value">Word-level sync</span>
<span class="hiw-detail-value">LLM-polished</span>
</div>
<div class="hiw-detail">
<span class="hiw-detail-label">Simultaneous Push</span>
@@ -576,7 +583,7 @@
<div class="hiw-cta-phone">
Or call in live: <strong>208-439-LUKE</strong>
</div>
<a href="https://ko-fi.com/lukemacneil" target="_blank" rel="noopener" class="hiw-cta-support">Support the show on Ko-fi</a>
<a href="https://ko-fi.com/lukemacneil" target="_blank" rel="noopener" class="hiw-cta-support">Support the Show</a>
</section>
<!-- Footer -->
@@ -602,9 +609,9 @@
<a href="https://youtube.com/lukemacneil" target="_blank" rel="noopener">YouTube</a>
</div>
</div>
<p class="footer-contact">Support the show: <a href="https://ko-fi.com/lukemacneil" target="_blank" rel="noopener">Ko-fi</a></p>
<p class="footer-contact"><a href="https://ko-fi.com/lukemacneil" target="_blank" rel="noopener">Support the Show</a></p>
<p class="footer-contact">Sales &amp; Collaboration: <a href="mailto:luke@macneilmediagroup.com">luke@macneilmediagroup.com</a></p>
<p>&copy; 2026 Luke at the Roost &middot; <a href="/privacy">Privacy Policy</a></p>
<p>&copy; 2026 Luke at the Roost &middot; <a href="/privacy">Privacy Policy</a> &middot; <a href="https://monitoring.macneilmediagroup.com/status/lukeattheroost" target="_blank" rel="noopener">System Status</a></p>
</footer>
</body>