Simplify architecture sections - diagrams first, remove ASCII art

README.md

@@ -105,70 +105,11 @@ We built a 5-agent pipeline that:
 
 The first half of our pipeline takes a massive YouTube video and identifies the **golden moments** worth sharing.
 
-### **The Agent Workflow**
-
-```
-┌─────────────────────────────────────────────────────────────────┐
-│           📥 PHASE 1: HOTSPOT DETECTION & CLIP CREATION         │
-└─────────────────────────────────────────────────────────────────┘
-
-YouTube URL (10min - 3hr video)
-    │
-    ▼
-┌─────────────────────────────────────────────────────────────────┐
-│  🔍 SCOUT AGENT                                                 │
-│  ─────────────────────────────────────────────────────────────  │
-│  Model: Gemini 2.0 Flash                                        │
-│  ─────────────────────────────────────────────────────────────  │
-│  • Downloads audio + fetches transcript                          │
-│  • Analyzes audio energy peaks (detects hype moments)            │
-│  • Semantic analysis of transcript for viral potential           │
-│  • Classifies video type (podcast/tutorial/vlog/gaming)          │
-│  • Outputs: Top 10-20 timestamp candidates with reasoning        │
-└──────────────────────────┬──────────────────────────────────────┘
-                           │
-                           ▼
-┌─────────────────────────────────────────────────────────────────┐
-│  ✅ VERIFIER AGENT                                              │
-│  ─────────────────────────────────────────────────────────────  │
-│  Model: Gemini 2.0 Flash + Vision API                           │
-│  ─────────────────────────────────────────────────────────────  │
-│  • Downloads 5-10s video segments around each hotspot            │
-│  • Uploads to Gemini Vision for visual quality analysis          │
-│  • Scores each clip: Visual Quality | Engagement | Shareability  │
-│  • Filters out low-quality, blurry, or boring segments           │
-│  • Outputs: Verified clips ranked by viral potential             │
-└──────────────────────────┬──────────────────────────────────────┘
-                           │
-                           ▼
-┌─────────────────────────────────────────────────────────────────┐
-│  🎬 DIRECTOR AGENT                                              │
-│  ─────────────────────────────────────────────────────────────  │
-│  Model: Gemini 2.0 Flash Lite                                   │
-│  ─────────────────────────────────────────────────────────────  │
-│  • Creates the edit plan from verified clips                     │
-│  • Determines optimal clip order for narrative flow              │
-│  • Sets duration per clip (8-15s each, 30-60s total)            │
-│  • Plans transition types and pacing                             │
-│  • Outputs: JSON edit plan with precise timestamps               │
-└──────────────────────────┬──────────────────────────────────────┘
-                           │
-                           ▼
-┌─────────────────────────────────────────────────────────────────┐
-│  🖐��� HANDS AGENT                                                 │
-│  ─────────────────────────────────────────────────────────────  │
-│  Tech: FFmpeg + Modal (Cloud Compute)                           │
-│  ─────────────────────────────────────────────────────────────  │
-│  • Extracts clips at exact timestamps (no re-download)           │
-│  • Concatenates with crossfade transitions                       │
-│  • Handles audio normalization                                   │
-│  • Outputs: Raw compiled video (30-60s landscape)                │
-└─────────────────────────────────────────────────────────────────┘
-                           │
-                           ▼
-                  🎥 RAW COMPILED CLIP
-                  (Ready for Production Polish)
-```
+<p align="center">
+  <img src="./resources/diagram1.png" width="700" alt="Architecture 1 - Clip Generation Pipeline">
+  <br>
+  <em>Scout → Verifier → Director → Hands: From YouTube URL to raw compiled clip</em>
+</p>
 
 ### **Gradio App - Clip Generation Interface**
 
@@ -178,80 +119,17 @@ YouTube URL (10min - 3hr video)
   <em>The Gradio interface for generating clips from YouTube URLs</em>
 </p>
 
-### **Architecture Block Diagram**
-
-<p align="center">
-  <img src="./resources/diagram1.png" width="700" alt="Architecture 1 - Clip Generation Pipeline">
-</p>
-
 ---
 
 ## 🏗️ **Architecture 2: Production Polish & Refinement**
 
 The second half takes that raw clip and transforms it into **viral-ready vertical content**.
 
-### **The Showrunner Pipeline**
-
-```
-┌─────────────────────────────────────────────────────────────────┐
-│           🎭 PHASE 2: PRODUCTION POLISH & REFINEMENT            │
-└─────────────────────────────────────────────────────────────────┘
-
-Raw Compiled Clip (30-60s landscape)
-    │
-    ▼
-┌─────────────────────────────────────────────────────────────────┐
-│  🎭 SHOWRUNNER AGENT - Step 1: Smart Crop                       │
-│  ─────────────────────────────────────────────────────────────  │
-│  Tech: Gemini Flash Lite + Qwen VL (Nebius)                     │
-│  ─────────────────────────────────────────────────────────────  │
-│  • Gemini detects scene change timestamps                        │
-│  • Qwen VL analyzes each key frame for subject position          │
-│  • Calculates optimal crop window (tracks faces/subjects)        │
-│  • Smooth interpolation between positions (no jarring jumps)     │
-│  • Renders 9:16 vertical with intelligent framing                │
-└──────────────────────────┬──────────────────────────────────────┘
-                           │
-                           ▼
-┌─────────────────────────────────────────────────────────────────┐
-│  🎭 SHOWRUNNER AGENT - Step 2: Intro Generation                 │
-│  ─────────────────────────────────────────────────────────────  │
-│  Tech: FLUX (Nebius) + ElevenLabs                               │
-│  ─────────────────────────────────────────────────────────────  │
-│  • Gemini writes content-aware hook script                       │
-│  • FLUX generates custom intro image (mood-matched)              │
-│  • ElevenLabs synthesizes professional voiceover                 │
-│  • Combines into 3-5s animated intro sequence                    │
-└──────────────────────────┬──────────────────────────���───────────┘
-                           │
-                           ▼
-┌─────────────────────────────────────────────────────────────────┐
-│  🎭 SHOWRUNNER AGENT - Step 3: Audio & Subtitles                │
-│  ─────────────────────────────────────────────────────────────  │
-│  Tech: WhisperX + FFmpeg                                        │
-│  ─────────────────────────────────────────────────────────────  │
-│  • WhisperX transcribes with word-level timestamps               │
-│  • Mood detection selects background music (hype/chill/suspense) │
-│  • Audio mixing: voice + music at optimal levels                 │
-│  • Burns stylized subtitles into video                           │
-└──────────────────────────┬──────────────────────────────────────┘
-                           │
-                           ▼
-┌─────────────────────────────────────────────────────────────────┐
-│  🎭 SHOWRUNNER AGENT - Step 4: Final Assembly                   │
-│  ─────────────────────────────────────────────────────────────  │
-│  Tech: FFmpeg + MoviePy                                         │
-│  ─────────────────────────────────────────────────────────────  │
-│  • Concatenates intro + main content                             │
-│  • Applies final color grading                                   │
-│  • Exports in TikTok/Reels/Shorts optimized format               │
-│  • Delivers download-ready MP4                                   │
-└─────────────────────────────────────────────────────────────────┘
-                           │
-                           ▼
-              📱 VIRAL-READY VERTICAL VIDEO
-              (9:16, subtitled, intro, music)
-```
+<p align="center">
+  <img src="./resources/diagram2.png" width="700" alt="Architecture 2 - Production Polish Pipeline">
+  <br>
+  <em>Showrunner: Smart Crop → Intro → Subtitles → Final Assembly</em>
+</p>
 
 ### **Gradio App - Production Studio Interface**
 
@@ -261,12 +139,6 @@ Raw Compiled Clip (30-60s landscape)
   <em>The Production Studio for adding polish to your clips</em>
 </p>
 
-### **Architecture Block Diagram**
-
-<p align="center">
-  <img src="./resources/diagram2.png" width="700" alt="Architecture 2 - Production Polish Pipeline">
-</p>
-
 ---
 
 ## 🤝 **Partner Technologies**