Searchable Video Library

from mixpeek import Mixpeek

client = Mixpeek(api_key="YOUR_API_KEY", namespace="video-library")

# 1. A bucket for the archive, and a collection that splits each video into scenes with transcripts and embeddings
bucket = client.buckets.create(
    bucket_name="video-archive",
    bucket_schema={
        "properties": {
            "video": {
                "type": "video",
            },
        },
    },
)
collection = client.collections.create(
    collection_name="video_library",
    source={"type": "bucket", "bucket_ids": [bucket["bucket_id"]]},
    feature_extractor={
        "feature_extractor_name": "multimodal_extractor",
        "version": "v1",
        "parameters": {
            "split_method": "scene",
            "run_transcription": True,
            "run_transcription_embedding": True,
        },
    },
)

# 2. Upload and process
client.buckets.upload(
    bucket["bucket_id"],
    blobs=[{"property": "video", "type": "video", "data": "s3://your-bucket/video-archive/all-hands-2026-06.mp4"}],
)
client.collections.trigger(collection["collection_id"])

# 3. Visual and transcript search fused with RRF, then a rerank against each scene transcript
retriever = client.retrievers.create(
    retriever_name="video-library-search",
    collection_identifiers=["video_library"],
    input_schema={
        "query": {
            "type": "text",
            "required": True,
        },
    },
    stages=[
        {
            "stage_name": "search",
            "stage_id": "feature_search",
            "parameters": {
                "searches": [
                    {
                        "feature_uri": "mixpeek://multimodal_extractor@v1/vertex_multimodal_embedding",
                        "query": {
                            "input_mode": "text",
                            "value": "{{INPUT.query}}",
                        },
                        "top_k": 50,
                    },
                    {
                        "feature_uri": "mixpeek://multimodal_extractor@v1/multilingual_e5_large_instruct_v1",
                        "query": {
                            "input_mode": "text",
                            "value": "{{INPUT.query}}",
                        },
                        "top_k": 50,
                    },
                ],
                "fusion": "rrf",
                "final_top_k": 50,
            },
        },
        {
            "stage_name": "rerank",
            "stage_id": "rerank",
            "parameters": {
                "inference_name": "BAAI__bge_reranker_v2_m3",
                "query": "{{INPUT.query}}",
                "document_field": "transcription",
                "top_k": 10,
            },
        },
    ],
)

# 4. Search
results = client.retrievers.execute(
    retriever["retriever_id"],
    inputs={
    "query": "product roadmap presentation Q3 goals",
},
)

for doc in results["documents"]:
    print(doc["source_object_id"], doc["start_time"], doc["end_time"], doc["score"])