{
  "id": "semantic_router",
  "title": "Semantic Routing",
  "url": "https://redis.io/docs/latest/develop/ai/redisvl/0.13.0/user_guide/semantic_router/",
  "summary": "",
  "content": "\n\nRedisVL provides a `SemanticRouter` interface to utilize Redis' built-in search \u0026 aggregation in order to perform\nKNN-style classification over a set of `Route` references to determine the best match.\n\nThis notebook will go over how to use Redis as a Semantic Router for your applications\n\n## Define the Routes\n\nBelow we define 3 different routes. One for `technology`, one for `sports`, and\nanother for `entertainment`. Now for this example, the goal here is\nsurely topic \"classification\". But you can create routes and references for\nalmost anything.\n\nEach route has a set of references that cover the \"semantic surface area\" of the\nroute. The incoming query from a user needs to be semantically similar to one or\nmore of the references in order to \"match\" on the route.\n\nAdditionally, each route has a `distance_threshold` which determines the maximum distance between the query and the reference for the query to be routed to the route. This value is unique to each route and uses Redis COSINE distance units (0-2], where lower values require stricter matching.\n\n\n```python\nfrom redisvl.extensions.router import Route\n\n# Define routes for the semantic router\ntechnology = Route(\n    name=\"technology\",\n    references=[\n        \"what are the latest advancements in AI?\",\n        \"tell me about the newest gadgets\",\n        \"what's trending in tech?\"\n    ],\n    metadata={\"category\": \"tech\", \"priority\": 1},\n    distance_threshold=0.71\n)\n\nsports = Route(\n    name=\"sports\",\n    references=[\n        \"who won the game last night?\",\n        \"tell me about the upcoming sports events\",\n        \"what's the latest in the world of sports?\",\n        \"sports\",\n        \"basketball and football\"\n    ],\n    metadata={\"category\": \"sports\", \"priority\": 2},\n    distance_threshold=0.72\n)\n\nentertainment = Route(\n    name=\"entertainment\",\n    references=[\n        \"what are the top movies right now?\",\n        \"who won the best actor award?\",\n        \"what's new in the entertainment industry?\"\n    ],\n    metadata={\"category\": \"entertainment\", \"priority\": 3},\n    distance_threshold=0.7\n)\n\n```\n\n## Initialize the SemanticRouter\n\n``SemanticRouter`` will automatically create an index within Redis upon initialization for the route references. By default, it uses the `HFTextVectorizer` to \ngenerate embeddings for each route reference.\n\n\n```python\nimport os\nfrom redisvl.extensions.router import SemanticRouter\nfrom redisvl.utils.vectorize import HFTextVectorizer\n\nos.environ[\"TOKENIZERS_PARALLELISM\"] = \"false\"\n\n# Initialize the SemanticRouter\nrouter = SemanticRouter(\n    name=\"topic-router\",\n    vectorizer=HFTextVectorizer(),\n    routes=[technology, sports, entertainment],\n    redis_url=\"redis://localhost:6379\",\n    overwrite=True # Blow away any other routing index with this name\n)\n```\n\n    /Users/tyler.hutcherson/Documents/AppliedAI/redis-vl-python/.venv/lib/python3.13/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n      from .autonotebook import tqdm as notebook_tqdm\n\n\n    13:03:49 sentence_transformers.SentenceTransformer INFO   Use pytorch device_name: mps\n    13:03:49 sentence_transformers.SentenceTransformer INFO   Load pretrained SentenceTransformer: sentence-transformers/all-mpnet-base-v2\n\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  6.31it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  7.02it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  8.21it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 54.33it/s]\n\n\n\n```python\n# look at the index specification created for the semantic router\n!rvl index info -i topic-router\n```\n\n    \n    \n    Index Information:\n    ╭──────────────────┬──────────────────┬──────────────────┬──────────────────┬──────────────────╮\n    │ Index Name       │ Storage Type     │ Prefixes         │ Index Options    │ Indexing         │\n    ├──────────────────┼──────────────────┼──────────────────┼──────────────────┼──────────────────┤\n    | topic-router     | HASH             | ['topic-router'] | []               | 0                |\n    ╰──────────────────┴──────────────────┴──────────────────┴──────────────────┴──────────────────╯\n    Index Fields:\n    ╭─────────────────┬─────────────────┬─────────────────┬─────────────────┬─────────────────┬─────────────────┬─────────────────┬─────────────────┬─────────────────┬─────────────────┬─────────────────╮\n    │ Name            │ Attribute       │ Type            │ Field Option    │ Option Value    │ Field Option    │ Option Value    │ Field Option    │ Option Value    │ Field Option    │ Option Value    │\n    ├─────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┼─────────────────┤\n    │ reference_id    │ reference_id    │ TAG             │ SEPARATOR       │ ,               │                 │                 │                 │                 │                 │                 │\n    │ route_name      │ route_name      │ TAG             │ SEPARATOR       │ ,               │                 │                 │                 │                 │                 │                 │\n    │ reference       │ reference       │ TEXT            │ WEIGHT          │ 1               │                 │                 │                 │                 │                 │                 │\n    │ vector          │ vector          │ VECTOR          │ algorithm       │ FLAT            │ data_type       │ FLOAT32         │ dim             │ 768             │ distance_metric │ COSINE          │\n    ╰─────────────────┴─────────────────┴─────────────────┴─────────────────┴─────────────────┴─────────────────┴─────────────────┴─────────────────┴─────────────────┴─────────────────┴─────────────────╯\n\n\n\n```python\nrouter._index.info()[\"num_docs\"]\n```\n\n\n\n\n    11\n\n\n\n## Simple routing\n\n\n```python\n# Query the router with a statement\nroute_match = router(\"Can you tell me about the latest in artificial intelligence?\")\nroute_match\n```\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  8.63it/s]\n\n\n\n\n\n    RouteMatch(name='technology', distance=0.419145941734)\n\n\n\n\n```python\n# Query the router with a statement and return a miss\nroute_match = router(\"are aliens real?\")\nroute_match\n```\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 11.71it/s]\n\n\n\n\n\n    RouteMatch(name=None, distance=None)\n\n\n\nWe can also route a statement to many routes and order them by distance:\n\n\n```python\n# Perform multi-class classification with route_many() -- toggle the max_k and the distance_threshold\nroute_matches = router.route_many(\"How is AI used in basketball?\", max_k=3)\nroute_matches\n```\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 12.12it/s]\n\n\n\n\n\n    [RouteMatch(name='technology', distance=0.556493639946),\n     RouteMatch(name='sports', distance=0.671060085297)]\n\n\n\n\n```python\n# Toggle the aggregation method -- note the different distances in the result\nfrom redisvl.extensions.router.schema import DistanceAggregationMethod\n\nroute_matches = router.route_many(\"How is AI used in basketball?\", aggregation_method=DistanceAggregationMethod.min, max_k=3)\nroute_matches\n```\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 56.69it/s]\n\n\n\n\n\n    [RouteMatch(name='technology', distance=0.556493639946),\n     RouteMatch(name='sports', distance=0.629264354706)]\n\n\n\nNote the different route match distances. This is because we used the `min` aggregation method instead of the default `avg` approach.\n\n## Update the routing config\n\n\n```python\nfrom redisvl.extensions.router import RoutingConfig\n\nrouter.update_routing_config(\n    RoutingConfig(aggregation_method=DistanceAggregationMethod.min, max_k=3)\n)\n```\n\n\n```python\nroute_matches = router.route_many(\"Lebron James\")\nroute_matches\n```\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 13.20it/s]\n\n\n\n\n\n    [RouteMatch(name='sports', distance=0.663253903389)]\n\n\n\n## Router serialization\n\n\n```python\nrouter.to_dict()\n```\n\n\n\n\n    {'name': 'topic-router',\n     'routes': [{'name': 'technology',\n       'references': ['what are the latest advancements in AI?',\n        'tell me about the newest gadgets',\n        \"what's trending in tech?\"],\n       'metadata': {'category': 'tech', 'priority': 1},\n       'distance_threshold': 0.71},\n      {'name': 'sports',\n       'references': ['who won the game last night?',\n        'tell me about the upcoming sports events',\n        \"what's the latest in the world of sports?\",\n        'sports',\n        'basketball and football'],\n       'metadata': {'category': 'sports', 'priority': 2},\n       'distance_threshold': 0.72},\n      {'name': 'entertainment',\n       'references': ['what are the top movies right now?',\n        'who won the best actor award?',\n        \"what's new in the entertainment industry?\"],\n       'metadata': {'category': 'entertainment', 'priority': 3},\n       'distance_threshold': 0.7}],\n     'vectorizer': {'type': 'hf',\n      'model': 'sentence-transformers/all-mpnet-base-v2'},\n     'routing_config': {'max_k': 3, 'aggregation_method': 'min'}}\n\n\n\n\n```python\nrouter2 = SemanticRouter.from_dict(router.to_dict(), redis_url=\"redis://localhost:6379\")\n\nassert router2.to_dict() == router.to_dict()\n```\n\n    13:03:54 sentence_transformers.SentenceTransformer INFO   Use pytorch device_name: mps\n    13:03:54 sentence_transformers.SentenceTransformer INFO   Load pretrained SentenceTransformer: sentence-transformers/all-mpnet-base-v2\n\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 53.91it/s]\n\n    13:03:54 redisvl.index.index INFO   Index already exists, not overwriting.\n\n\n    \n\n\n\n```python\nrouter.to_yaml(\"router.yaml\", overwrite=True)\n```\n\n\n```python\nrouter3 = SemanticRouter.from_yaml(\"router.yaml\", redis_url=\"redis://localhost:6379\")\n\nassert router3.to_dict() == router2.to_dict() == router.to_dict()\n```\n\n    13:03:54 sentence_transformers.SentenceTransformer INFO   Use pytorch device_name: mps\n    13:03:54 sentence_transformers.SentenceTransformer INFO   Load pretrained SentenceTransformer: sentence-transformers/all-mpnet-base-v2\n\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 51.94it/s]\n\n    13:03:55 redisvl.index.index INFO   Index already exists, not overwriting.\n\n\n    \n\n\n## Add route references\n\n\n```python\nrouter.add_route_references(route_name=\"technology\", references=[\"latest AI trends\", \"new tech gadgets\"])\n```\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  8.12it/s]\n\n\n\n\n\n    ['topic-router:technology:f243fb2d073774e81c7815247cb3013794e6225df3cbe3769cee8c6cefaca777',\n     'topic-router:technology:7e4bca5853c1c3298b4d001de13c3c7a79a6e0f134f81acc2e7cddbd6845961f']\n\n\n\n## Get route references\n\n\n```python\n# by route name\nrefs = router.get_route_references(route_name=\"technology\")\nrefs\n```\n\n\n\n\n    [{'id': 'topic-router:technology:85cc73a1437df27caa2f075a29c497e5a2e532023fbb75378aedbae80779ab37',\n      'reference_id': '85cc73a1437df27caa2f075a29c497e5a2e532023fbb75378aedbae80779ab37',\n      'route_name': 'technology',\n      'reference': 'tell me about the newest gadgets'},\n     {'id': 'topic-router:technology:851f51cce5a9ccfbbcb66993908be6b7871479af3e3a4b139ad292a1bf7e0676',\n      'reference_id': '851f51cce5a9ccfbbcb66993908be6b7871479af3e3a4b139ad292a1bf7e0676',\n      'route_name': 'technology',\n      'reference': 'what are the latest advancements in AI?'},\n     {'id': 'topic-router:technology:f243fb2d073774e81c7815247cb3013794e6225df3cbe3769cee8c6cefaca777',\n      'reference_id': 'f243fb2d073774e81c7815247cb3013794e6225df3cbe3769cee8c6cefaca777',\n      'route_name': 'technology',\n      'reference': 'latest AI trends'},\n     {'id': 'topic-router:technology:7e4bca5853c1c3298b4d001de13c3c7a79a6e0f134f81acc2e7cddbd6845961f',\n      'reference_id': '7e4bca5853c1c3298b4d001de13c3c7a79a6e0f134f81acc2e7cddbd6845961f',\n      'route_name': 'technology',\n      'reference': 'new tech gadgets'},\n     {'id': 'topic-router:technology:149a9c9919c58534aa0f369e85ad95ba7f00aa0513e0f81e2aff2ea4a717b0e0',\n      'reference_id': '149a9c9919c58534aa0f369e85ad95ba7f00aa0513e0f81e2aff2ea4a717b0e0',\n      'route_name': 'technology',\n      'reference': \"what's trending in tech?\"}]\n\n\n\n\n```python\n# by reference id\nrefs = router.get_route_references(reference_ids=[refs[0][\"reference_id\"]])\nrefs\n```\n\n\n\n\n    [{'id': 'topic-router:technology:85cc73a1437df27caa2f075a29c497e5a2e532023fbb75378aedbae80779ab37',\n      'reference_id': '85cc73a1437df27caa2f075a29c497e5a2e532023fbb75378aedbae80779ab37',\n      'route_name': 'technology',\n      'reference': 'tell me about the newest gadgets'}]\n\n\n\n## Delete route references\n\n\n```python\n# by route name\ndeleted_count = router.delete_route_references(route_name=\"sports\")\ndeleted_count\n```\n\n\n\n\n    5\n\n\n\n\n```python\n# by id\ndeleted_count = router.delete_route_references(reference_ids=[refs[0][\"reference_id\"]])\ndeleted_count\n```\n\n\n\n\n    1\n\n\n\n## Clean up the router\n\n\n```python\n# Use clear to flush all routes from the index\nrouter.clear()\n```\n\n\n```python\n# Use delete to clear the index and remove it completely\nrouter.delete()\n```\n",
  "tags": [],
  "last_updated": "2026-04-21T14:39:33+02:00"
}