search_proximity - Augent

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

Example

Request:

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

Response:

{
  "query": "'burnout' within 30 words of 'boundaries'",
  "match_count": 2,
  "matches": [
    {
      "timestamp": "3:42",
      "snippet": "...burnout is real and setting boundaries early is the only way to...",
      "distance": 8
    }
  ],
  "model_used": "tiny"
}

Parameters

Parameter	Required	Default	Description
`audio_path`	Yes	—	Path to the audio/video file
`keyword1`	Yes	—	Primary keyword to find
`keyword2`	Yes	—	Secondary keyword that must appear nearby
`max_distance`	No	`30`	Maximum number of words between keywords
`model_size`	No	`tiny`	Whisper model size for transcription
`output`	No	—	File path to save results (`.csv` or `.xlsx`)

Notes

Example use cases: “problem” near “solution”, “feature” near “deadline”, “competitor” near “advantage”