search_proximity

Find where one keyword appears near another keyword in audio. Useful for finding contextual discussions.

Example

Request:


{
  "audio_path": "/Users/you/Downloads/interview.webm",
  "keyword1": "startup",
  "keyword2": "funding",
  "max_distance": 30
}

Response:


{
  "query": "'startup' within 30 words of 'funding'",
  "match_count": 2,
  "matches": [
    {
      "timestamp": "3:42",
      "snippet": "...when the startup finally got its series A funding...",
      "distance": 8
    }
  ],
  "model_used": "tiny"
}

Notes

This finds contextual discussions — not just individual keywords, but keywords appearing in the same thought or sentence.

Example use cases: “problem” near “solution”, “price” near “expensive”, “competitor” near “advantage”