Enhanced

Feature Extraction

Multi-tier feature extraction that decomposes content into searchable components: embeddings, transcripts, detected objects, OCR text, scene boundaries, and more. The foundation for all downstream retrieval and analysis.

video

image

audio

text

Multi-Tier

142.0K runs

Deploy Recipe

"Find meeting recordings where someone discusses quarterly roadmap near a whiteboard"

Why This Matters

Raw media is unsearchable. Feature extraction transforms video, images, and audio into structured, queryable representations that power every other recipe.

import requests

API_URL = "https://api.mixpeek.com"
headers = {"Authorization": "Bearer YOUR_API_KEY", "X-Namespace": "your-namespace"}

# Create collection with feature extractor
collection = requests.post(f"{API_URL}/v1/collections", headers=headers, json={
    "collection_name": "enriched_media",
    "source": {"type": "bucket", "bucket_id": "raw-media"},
    "feature_extractor": {
        "feature_extractor_name": "multimodal_extractor",
        "version": "v1",
        "input_mappings": {"video": "source_video"},
        "parameters": {"enable_transcription": True}
    }
}).json()

# Index content - extraction happens automatically
requests.post(f"{API_URL}/v1/buckets/raw-media/objects", headers=headers, json={
    "blobs": [{"property": "source_video", "url": "s3://bucket/meeting.mp4"}]
})

# Search across all extracted features
results = requests.post(
    f"{API_URL}/v1/retrievers/enriched-search/execute",
    headers=headers,
    json={"query": {"text": "quarterly roadmap discussion"}}
).json()

# Access extracted features directly
for doc in results["documents"]:
    print(f"Transcript: {doc.get('transcript', '')[:200]}...")
    print(f"Feature URI: {doc['feature_address']}")

Feature Extractors

Image Embedding

Generate visual embeddings for similarity search and clustering

752K runs

Video Embedding

Generate vector embeddings for video content

610K runs

Audio Transcription

Transcribe audio content to text

450K runs

Text Embedding

Extract semantic embeddings from documents, transcripts and text content

827K runs

Object Detection

Identify and locate objects within images with bounding boxes

631K runs

+2 more extractors

Retriever Stages

feature search

Search and filter documents by vector similarity using feature embeddings

filter

attribute filter

Filter documents by metadata attribute values using boolean logic