HALOWERK buerowerk — Absolute Date Token Extraction for Meeting Transcripts

Use this endpoint when you have structured meeting transcript data (with timestamps and speaker labels) and need to extract and resolve date references into absolute tokens. It is specifically designed for the meeting context and handles both the flat segments format and the richer transcript_package format with media metadata. Prefer this over general NLP date extractors when your input is already segmented with timing information.

• Segments array is empty or missing required fields (start, end, text) — returns 400 validation error
• start or end timestamps exceed allowed maximums (14400s for segments, 1801s for cues) — returns validation error
• text fields exceed 3000 character limit per segment — returns 400 error
• No date references present in transcript — returns empty token list
• Both segments and transcript_package omitted — returns 400 missing input error
• Payment not provided or insufficient — returns 402 Payment Required