cueseg: topic chapters from a transcript, and a worked pipeline in the readme

parity/cases/cueseg/one-cue.ndjson

@@ -0,0 +1 @@
+{"start": 0.0, "end": 4.0, "text": "the mill stood at the northern end"}