Add Whisper pacing and Channel History features

candidates.json

@@ -70,6 +70,11 @@
       "type": "silence",
       "score": 0.6
     },
+    {
+      "timestamp": 275.88,
+      "type": "transcript_boundary",
+      "score": 0.7
+    },
     {
       "timestamp": 286.89,
       "type": "silence",

channel_history.json

@@ -0,0 +1,10 @@
+[
+  {
+    "timestamp": "2026-06-11 11:30",
+    "video_id": "4Rq-LY16WxM",
+    "video_path": "flashgard.mp4",
+    "duration": "9m 27s",
+    "placements": 3,
+    "is_prediction": false
+  }
+]

dashboard.py

@@ -902,6 +902,22 @@ with st.expander("🔬 View All Analyzed Moments (Advanced)"):
         st.dataframe(df_show, width='stretch', hide_index=True)
 
 
+# ═══════════════════════════════════════════════════════════════
+#  PAST ANALYSES (CHANNEL HISTORY)
+# ═══════════════════════════════════════════════════════════════
+history_data = load_json("channel_history.json")
+if history_data:
+    st.markdown('<div class="gradient-divider"></div>', unsafe_allow_html=True)
+    with st.expander("📚 Past Analyses (Channel History)"):
+        # Reverse the list so newest is on top
+        history_data = history_data[::-1]
+        df_history = pd.DataFrame(history_data)
+        df_history = df_history[["timestamp", "video_path", "duration", "placements", "is_prediction"]]
+        df_history.columns = ["Date", "Video", "Duration", "Spots Found", "Predicted?"]
+        df_history["Predicted?"] = df_history["Predicted?"].apply(lambda x: "🔮 Yes" if x else "✅ No")
+        st.dataframe(df_history, width='stretch', hide_index=True)
+
+
 # ═══════════════════════════════════════════════════════════════
 #  FOOTER
 # ═══════════════════════════════════════════════════════════════

pipeline.py

@@ -5,9 +5,24 @@ pipeline.py — Runs full pipeline as functions (no subprocess needed on Render)
 import json
 import os
 import tempfile
+import datetime
 import numpy as np
 import pandas as pd
 
+def save_to_history(entry):
+    """Save an analysis result to channel history."""
+    history_file = "channel_history.json"
+    history = []
+    if os.path.exists(history_file):
+        with open(history_file, "r") as f:
+            try:
+                history = json.load(f)
+            except:
+                pass
+    history.append(entry)
+    with open(history_file, "w") as f:
+        json.dump(history, f, indent=2)
+
 
 # ── Component 1: Candidate Generator ──────────────────────────
 
@@ -420,6 +435,17 @@ def run_full_pipeline(video_path, video_id, creds, progress_callback=None):
     update("🎯 Step 4/4 — Generating final recommendations...")
     result = run_component4()
 
+    # Save to history
+    entry = {
+        "timestamp": datetime.datetime.now().strftime("%Y-%m-%d %H:%M"),
+        "video_id": video_id,
+        "video_path": os.path.basename(video_path),
+        "duration": result.get("video_duration_formatted", "0m 0s"),
+        "placements": result.get("total_placements_recommended", 0),
+        "is_prediction": False
+    }
+    save_to_history(entry)
+
     update("✅ Analysis complete!")
     return result
 
@@ -450,5 +476,16 @@ def run_full_pipeline_offline(video_path, progress_callback=None):
     update("🎯 Step 4/4 — Generating final predictions...")
     result = run_component4()
 
+    # Save to history
+    entry = {
+        "timestamp": datetime.datetime.now().strftime("%Y-%m-%d %H:%M"),
+        "video_id": None,
+        "video_path": os.path.basename(video_path),
+        "duration": result.get("video_duration_formatted", "0m 0s"),
+        "placements": result.get("total_placements_recommended", 0),
+        "is_prediction": True
+    }
+    save_to_history(entry)
+
     update("✅ Prediction complete!")
     return result