feat(infra): transcribe a stored video by id and write its captions

apps/web/__tests__/fixtures/developer-captions-golden.vtt

@@ -0,0 +1,41 @@
+WEBVTT
+
+00:00:00.240 --> 00:00:01.240
+Welcome back.
+
+00:00:01.600 --> 00:00:06.320
+Today we are walking through the
+quarterly numbers for the whole company,
+
+00:00:06.400 --> 00:00:13.120
+region by region, and then we will look at
+what changed in the product since the last
+
+00:00:13.200 --> 00:00:15.920
+time we all met in this room.
+
+00:00:18.400 --> 00:00:21.928
+We shipped 1,250 units in
+Q3 2026 for $3.5 million.
+
+00:00:23.200 --> 00:00:26.400
+And yes I know that is a lot to take
+in all at once so bear with me please.
+
+00:00:26.400 --> 00:00:27.520
+Let me think
+
+00:00:31.200 --> 00:00:33.120
+about that for a second.
+
+00:00:34.400 --> 00:00:37.520
+Is 5 < 7 && 9 > 2?
+
+00:00:37.600 --> 00:00:40.320
+Tom & Jerry said <b>yes</b> --> maybe.
+
+00:00:41.600 --> 00:00:42.600
+Okay.
+
+00:00:43.200 --> 00:00:44.800
+Line break inside end.