radio: attach caller names to Q&A pairs from transcript intros

QAPair gets caller_name and caller_role fields populated by a new
attach_caller_names(pairs, transcript_segments) helper. For each pair,
finds the active opening intro at the question_start time (8s forward
tolerance, no backward limit — a caller's call can run for 10+ minutes
and the intro happens once at the start) and attaches the speaker name.

Validation on 9-episode test set:
  19/19 Q&A pairs (100%) now have caller names attached.

Examples of corrections from oracle attribution:
  2018-s10e18 @ 73:36  Christopher (was misattributed to "Tara")
  2015-s7e19 @ 35:45   William     (was misattributed to "Tara")
  2010-05-08-hr1       Jackie x3, Bruce
  2012-03-10-hr1       Adam x2
  2016-s8e43           John, Doug
  2017-s9e30           Tom, Denise x3, Charlie

speaker_oracle.py: adds speaker_at(time, intros) helper used both by the
existing resolve_speakers() and the new caller-name attachment. Also
adds the "let's fit/bring/put X in/on" intro pattern variant (caught
Charlie at 70:21 in 2017-s9e30 that "talk to X" missed).

download_full_archive.py: SSH keepalive every 30s + per-file retry-on-
failure (up to 3 attempts with reconnect). Earlier run hung on a dead
connection at file 109 of 589 with no recovery; restarted run is now
running at ~10 MB/s vs ~2-3 MB/s before.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

projects/radio-show/audio-processor/src/qa_extractor.py

@@ -78,6 +78,10 @@ class QAPair:
     answer_text: str
     topic: Optional[str] = None
     topic_tags: list[str] = field(default_factory=list)
+    # Caller identification — populated by attach_caller_names() when a
+    # transcript-side speaker intro precedes the question turn.
+    caller_name: Optional[str] = None
+    caller_role: Optional[str] = None  # "caller" / "guest" / "fillin"
 
     def to_dict(self) -> dict:
         return {
@@ -89,6 +93,8 @@ class QAPair:
             "answer_text": self.answer_text,
             "topic": self.topic,
             "topic_tags": self.topic_tags,
+            "caller_name": self.caller_name,
+            "caller_role": self.caller_role,
         }
 
     def clip_start(self, padding: float = 1.5) -> float:
@@ -101,6 +107,24 @@ class QAPair:
         return self.answer_end - self.question_start
 
 
+def attach_caller_names(pairs: list[QAPair],
+                        transcript_segments: list[dict]) -> list[QAPair]:
+    """Attach caller names to Q&A pairs using transcript-side speaker intros.
+
+    For each pair, looks up the active intro at the question_start time and
+    populates caller_name / caller_role. Mutates the input pairs and returns
+    the list for chaining.
+    """
+    from .speaker_oracle import extract_intros, speaker_at
+    intros = extract_intros(transcript_segments)
+    for pair in pairs:
+        intro = speaker_at(pair.question_start, intros)
+        if intro is not None:
+            pair.caller_name = intro.name
+            pair.caller_role = intro.role_hint
+    return pairs
+
+
 def extract_qa_pairs(diarized_segments: list[dict]) -> list[QAPair]:
     """
     Extract caller Q&A pairs from diarized transcript segments.