{
  "id": "getting_started",
  "title": "Getting Started with RedisVL",
  "url": "https://redis.io/docs/latest/develop/ai/redisvl/0.11.0/user_guide/getting_started/",
  "summary": "",
  "content": "\n`redisvl` is a versatile Python library with an integrated CLI, designed to enhance AI applications using Redis. This guide will walk you through the following steps:\n\n1. Defining an `IndexSchema`\n2. Preparing a sample dataset\n3. Creating a `SearchIndex` object\n4. Testing `rvl` CLI functionality\n5. Loading the sample data\n6. Building `VectorQuery` objects and executing searches\n7. Updating a `SearchIndex` object\n\n...and more!\n\nPrerequisites:\n- Ensure `redisvl` is installed in your Python environment.\n- Have a running instance of [Redis Stack](https://redis.io/docs/install/install-stack/) or [Redis Cloud](https://redis.io/cloud).\n\n_____\n\n## Define an `IndexSchema`\n\nThe `IndexSchema` maintains crucial **index configuration** and **field definitions** to\nenable search with Redis. For ease of use, the schema can be constructed from a\npython dictionary or yaml file.\n\n### Example Schema Creation\nConsider a dataset with user information, including `job`, `age`, `credit_score`,\nand a 3-dimensional `user_embedding` vector.\n\nYou must also decide on a Redis index name and key prefix to use for this\ndataset. Below are example schema definitions in both YAML and Dict format.\n\n**YAML Definition:**\n\n```yaml\nversion: '0.1.0'\n\nindex:\n  name: user_simple\n  prefix: user_simple_docs\n\nfields:\n    - name: user\n      type: tag\n    - name: credit_score\n      type: tag\n    - name: job\n      type: text\n    - name: age\n      type: numeric\n    - name: user_embedding\n      type: vector\n      attrs:\n        algorithm: flat\n        dims: 3\n        distance_metric: cosine\n        datatype: float32\n```\nStore this in a local file, such as `schema.yaml`, for RedisVL usage.\n\n**Python Dictionary:**\n\n\n```python\nschema = {\n    \"index\": {\n        \"name\": \"user_simple\",\n        \"prefix\": \"user_simple_docs\",\n    },\n    \"fields\": [\n        {\"name\": \"user\", \"type\": \"tag\"},\n        {\"name\": \"credit_score\", \"type\": \"tag\"},\n        {\"name\": \"job\", \"type\": \"text\"},\n        {\"name\": \"age\", \"type\": \"numeric\"},\n        {\n            \"name\": \"user_embedding\",\n            \"type\": \"vector\",\n            \"attrs\": {\n                \"dims\": 3,\n                \"distance_metric\": \"cosine\",\n                \"algorithm\": \"flat\",\n                \"datatype\": \"float32\"\n            }\n        }\n    ]\n}\n```\n\n## Sample Dataset Preparation\n\nBelow, create a mock dataset with `user`, `job`, `age`, `credit_score`, and\n`user_embedding` fields. The `user_embedding` vectors are synthetic examples\nfor demonstration purposes.\n\nFor more information on creating real-world embeddings, refer to this\n[article](https://mlops.community/vector-similarity-search-from-basics-to-production/).\n\n\n```python\nimport numpy as np\n\n\ndata = [\n    {\n        'user': 'john',\n        'age': 1,\n        'job': 'engineer',\n        'credit_score': 'high',\n        'user_embedding': np.array([0.1, 0.1, 0.5], dtype=np.float32).tobytes()\n    },\n    {\n        'user': 'mary',\n        'age': 2,\n        'job': 'doctor',\n        'credit_score': 'low',\n        'user_embedding': np.array([0.1, 0.1, 0.5], dtype=np.float32).tobytes()\n    },\n    {\n        'user': 'joe',\n        'age': 3,\n        'job': 'dentist',\n        'credit_score': 'medium',\n        'user_embedding': np.array([0.9, 0.9, 0.1], dtype=np.float32).tobytes()\n    }\n]\n```\n\nAs seen above, the sample `user_embedding` vectors are converted into bytes. Using the `NumPy`, this is fairly trivial.\n\n## Create a `SearchIndex`\n\nWith the schema and sample dataset ready, create a `SearchIndex`.\n\n### Bring your own Redis connection instance\n\nThis is ideal in scenarios where you have custom settings on the connection instance or if your application will share a connection pool:\n\n\n```python\nfrom redisvl.index import SearchIndex\nfrom redis import Redis\n\nclient = Redis.from_url(\"redis://localhost:6379\")\nindex = SearchIndex.from_dict(schema, redis_client=client, validate_on_load=True)\n```\n\n### Let the index manage the connection instance\n\nThis is ideal for simple cases:\n\n\n```python\nindex = SearchIndex.from_dict(schema, redis_url=\"redis://localhost:6379\", validate_on_load=True)\n\n# If you don't specify a client or Redis URL, the index will attempt to\n# connect to Redis at the default address \"redis://localhost:6379\".\n```\n\n### Create the index\n\nNow that we are connected to Redis, we need to run the create command.\n\n\n```python\nindex.create(overwrite=True)\n```\n\n    13:00:22 redisvl.index.index INFO   Index already exists, overwriting.\n\n\nNote that at this point, the index has no entries. Data loading follows.\n\n## Inspect with the `rvl` CLI\nUse the `rvl` CLI to inspect the created index and its fields:\n\n\n```python\n!rvl index listall\n```\n\n    13:00:24 [RedisVL] INFO   Indices:\n    13:00:24 [RedisVL] INFO   1. user_simple\n\n\n\n```python\n!rvl index info -i user_simple\n```\n\n    \n    \n    Index Information:\n    ╭──────────────────────┬──────────────────────┬──────────────────────┬──────────────────────┬──────────────────────╮\n    │ Index Name           │ Storage Type         │ Prefixes             │ Index Options        │ Indexing             │\n    ├──────────────────────┼──────────────────────┼──────────────────────┼──────────────────────┼──────────────────────┤\n    | user_simple          | HASH                 | ['user_simple_docs'] | []                   | 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    │ user            │ user            │ TAG             │ SEPARATOR       │ ,               │                 │                 │                 │                 │                 │                 │\n    │ credit_score    │ credit_score    │ TAG             │ SEPARATOR       │ ,               │                 │                 │                 │                 │                 │                 │\n    │ job             │ job             │ TEXT            │ WEIGHT          │ 1               │                 │                 │                 │                 │                 │                 │\n    │ age             │ age             │ NUMERIC         │                 │                 │                 │                 │                 │                 │                 │                 │\n    │ user_embedding  │ user_embedding  │ VECTOR          │ algorithm       │ FLAT            │ data_type       │ FLOAT32         │ dim             │ 3               │ distance_metric │ COSINE          │\n    ╰─────────────────┴─────────────────┴─────────────────┴─────────────────┴─────────────────┴─────────────────┴─────────────────┴─────────────────┴─────────────────┴─────────────────┴─────────────────╯\n\n\n## Load Data to `SearchIndex`\n\nLoad the sample dataset to Redis.\n\n### Validate data entries on load\nRedisVL uses pydantic validation under the hood to ensure loaded data is valid and confirms to your schema. This setting is optional and can be configured in the `SearchIndex` class.\n\n\n```python\nkeys = index.load(data)\n\nprint(keys)\n```\n\n    ['user_simple_docs:01JY4J4Y08GFY10VMB9D4YDMZQ', 'user_simple_docs:01JY4J4Y0AY2MKJ24QXQS2Q2YS', 'user_simple_docs:01JY4J4Y0A9GFF2XG1R81EFD4Z']\n\n\nBy default, `load` will create a unique Redis key as a combination of the index key `prefix` and a random ULID. You can also customize the key by providing direct keys or pointing to a specified `id_field` on load.\n\n### Load INVALID data\nThis will raise a `SchemaValidationError` if `validate_on_load` is set to true in the `SearchIndex` class.\n\n\n```python\n# NBVAL_SKIP\n\ntry:\n    keys = index.load([{\"user_embedding\": True}])\nexcept Exception as e:\n    print(str(e))\n```\n\n    13:00:27 redisvl.index.index ERROR   Data validation failed during load operation\n    Schema validation failed for object at index 0. Field 'user_embedding' expects bytes (vector data), but got boolean value 'True'. If this should be a vector field, provide a list of numbers or bytes. If this should be a different field type, check your schema definition.\n    Object data: {\n      \"user_embedding\": true\n    }\n    Hint: Check that your data types match the schema field definitions. Use index.schema.fields to view expected field types.\n\n\n### Upsert the index with new data\nUpsert data by using the `load` method again:\n\n\n```python\n# Add more data\nnew_data = [{\n    'user': 'tyler',\n    'age': 9,\n    'job': 'engineer',\n    'credit_score': 'high',\n    'user_embedding': np.array([0.1, 0.3, 0.5], dtype=np.float32).tobytes()\n}]\nkeys = index.load(new_data)\n\nprint(keys)\n```\n\n    ['user_simple_docs:01JY4J4Y0N4CNR9Y6R67MMVG7Q']\n\n\n## Creating `VectorQuery` Objects\n\nNext we will create a vector query object for our newly populated index. This example will use a simple vector to demonstrate how vector similarity works. Vectors in production will likely be much larger than 3 floats and often require Machine Learning models (i.e. Huggingface sentence transformers) or an embeddings API (Cohere, OpenAI). `redisvl` provides a set of [Vectorizers]() to assist in vector creation.\n\n\n```python\nfrom redisvl.query import VectorQuery\nfrom jupyterutils import result_print\n\nquery = VectorQuery(\n    vector=[0.1, 0.1, 0.5],\n    vector_field_name=\"user_embedding\",\n    return_fields=[\"user\", \"age\", \"job\", \"credit_score\", \"vector_distance\"],\n    num_results=3\n)\n```\n\n### Executing queries\nWith our `VectorQuery` object defined above, we can execute the query over the `SearchIndex` using the `query` method.\n\n\n```python\nresults = index.query(query)\nresult_print(results)\n```\n\n\n\u003ctable\u003e\u003ctr\u003e\u003cth\u003evector_distance\u003c/th\u003e\u003cth\u003euser\u003c/th\u003e\u003cth\u003eage\u003c/th\u003e\u003cth\u003ejob\u003c/th\u003e\u003cth\u003ecredit_score\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e0\u003c/td\u003e\u003ctd\u003ejohn\u003c/td\u003e\u003ctd\u003e1\u003c/td\u003e\u003ctd\u003eengineer\u003c/td\u003e\u003ctd\u003ehigh\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e0\u003c/td\u003e\u003ctd\u003emary\u003c/td\u003e\u003ctd\u003e2\u003c/td\u003e\u003ctd\u003edoctor\u003c/td\u003e\u003ctd\u003elow\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e0\u003c/td\u003e\u003ctd\u003ejohn\u003c/td\u003e\u003ctd\u003e1\u003c/td\u003e\u003ctd\u003eengineer\u003c/td\u003e\u003ctd\u003ehigh\u003c/td\u003e\u003c/tr\u003e\u003c/table\u003e\n\n\n## Using an Asynchronous Redis Client\n\nThe `AsyncSearchIndex` class along with an async Redis python client allows for queries, index creation, and data loading to be done asynchronously. This is the\nrecommended route for working with `redisvl` in production-like settings.\n\n\n```python\nschema\n```\n\n\n\n\n    {'index': {'name': 'user_simple', 'prefix': 'user_simple_docs'},\n     'fields': [{'name': 'user', 'type': 'tag'},\n      {'name': 'credit_score', 'type': 'tag'},\n      {'name': 'job', 'type': 'text'},\n      {'name': 'age', 'type': 'numeric'},\n      {'name': 'user_embedding',\n       'type': 'vector',\n       'attrs': {'dims': 3,\n        'distance_metric': 'cosine',\n        'algorithm': 'flat',\n        'datatype': 'float32'}}]}\n\n\n\n\n```python\nfrom redisvl.index import AsyncSearchIndex\nfrom redis.asyncio import Redis\n\nclient = Redis.from_url(\"redis://localhost:6379\")\nindex = AsyncSearchIndex.from_dict(schema, redis_client=client)\n```\n\n\n```python\n# execute the vector query async\nresults = await index.query(query)\nresult_print(results)\n```\n\n\n\u003ctable\u003e\u003ctr\u003e\u003cth\u003evector_distance\u003c/th\u003e\u003cth\u003euser\u003c/th\u003e\u003cth\u003eage\u003c/th\u003e\u003cth\u003ejob\u003c/th\u003e\u003cth\u003ecredit_score\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e0\u003c/td\u003e\u003ctd\u003ejohn\u003c/td\u003e\u003ctd\u003e1\u003c/td\u003e\u003ctd\u003eengineer\u003c/td\u003e\u003ctd\u003ehigh\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e0\u003c/td\u003e\u003ctd\u003emary\u003c/td\u003e\u003ctd\u003e2\u003c/td\u003e\u003ctd\u003edoctor\u003c/td\u003e\u003ctd\u003elow\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e0\u003c/td\u003e\u003ctd\u003ejohn\u003c/td\u003e\u003ctd\u003e1\u003c/td\u003e\u003ctd\u003eengineer\u003c/td\u003e\u003ctd\u003ehigh\u003c/td\u003e\u003c/tr\u003e\u003c/table\u003e\n\n\n## Updating a schema\nIn some scenarios, it makes sense to update the index schema. With Redis and `redisvl`, this is easy because Redis can keep the underlying data in place while you change or make updates to the index configuration.\n\nSo for our scenario, let's imagine we want to reindex this data in 2 ways:\n- by using a `Tag` type for `job` field instead of `Text`\n- by using an `hnsw` vector index for the `user_embedding` field instead of a `flat` vector index\n\n\n```python\n# Modify this schema to have what we want\n\nindex.schema.remove_field(\"job\")\nindex.schema.remove_field(\"user_embedding\")\nindex.schema.add_fields([\n    {\"name\": \"job\", \"type\": \"tag\"},\n    {\n        \"name\": \"user_embedding\",\n        \"type\": \"vector\",\n        \"attrs\": {\n            \"dims\": 3,\n            \"distance_metric\": \"cosine\",\n            \"algorithm\": \"hnsw\",\n            \"datatype\": \"float32\"\n        }\n    }\n])\n```\n\n\n```python\n# Run the index update but keep underlying data in place\nawait index.create(overwrite=True, drop=False)\n```\n\n    13:00:27 redisvl.index.index INFO   Index already exists, overwriting.\n\n\n\n```python\n# Execute the vector query async\nresults = await index.query(query)\nresult_print(results)\n```\n\n\n\u003ctable\u003e\u003ctr\u003e\u003cth\u003evector_distance\u003c/th\u003e\u003cth\u003euser\u003c/th\u003e\u003cth\u003eage\u003c/th\u003e\u003cth\u003ejob\u003c/th\u003e\u003cth\u003ecredit_score\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e0\u003c/td\u003e\u003ctd\u003emary\u003c/td\u003e\u003ctd\u003e2\u003c/td\u003e\u003ctd\u003edoctor\u003c/td\u003e\u003ctd\u003elow\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e0\u003c/td\u003e\u003ctd\u003ejohn\u003c/td\u003e\u003ctd\u003e1\u003c/td\u003e\u003ctd\u003eengineer\u003c/td\u003e\u003ctd\u003ehigh\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd\u003e0\u003c/td\u003e\u003ctd\u003ejohn\u003c/td\u003e\u003ctd\u003e1\u003c/td\u003e\u003ctd\u003eengineer\u003c/td\u003e\u003ctd\u003ehigh\u003c/td\u003e\u003c/tr\u003e\u003c/table\u003e\n\n\n## Check Index Stats\nUse the `rvl` CLI to check the stats for the index:\n\n\n```python\n!rvl stats -i user_simple\n```\n\n    \n    Statistics:\n    ╭─────────────────────────────┬────────────╮\n    │ Stat Key                    │ Value      │\n    ├─────────────────────────────┼────────────┤\n    │ num_docs                    │ 10         │\n    │ num_terms                   │ 0          │\n    │ max_doc_id                  │ 10         │\n    │ num_records                 │ 50         │\n    │ percent_indexed             │ 1          │\n    │ hash_indexing_failures      │ 0          │\n    │ number_of_uses              │ 2          │\n    │ bytes_per_record_avg        │ 19.5200004 │\n    │ doc_table_size_mb           │ 0.00105857 │\n    │ inverted_sz_mb              │ 9.30786132 │\n    │ key_table_size_mb           │ 4.70161437 │\n    │ offset_bits_per_record_avg  │ nan        │\n    │ offset_vectors_sz_mb        │ 0          │\n    │ offsets_per_term_avg        │ 0          │\n    │ records_per_doc_avg         │ 5          │\n    │ sortable_values_size_mb     │ 0          │\n    │ total_indexing_time         │ 0.16899999 │\n    │ total_inverted_index_blocks │ 11         │\n    │ vector_index_sz_mb          │ 0.23619842 │\n    ╰─────────────────────────────┴────────────╯\n\n\n## Cleanup\n\nBelow we will clean up after our work. First, you can flush all data from Redis associated with the index by\nusing the `.clear()` method. This will leave the secondary index in place for future insertions or updates.\n\nBut if you want to clean up everything, including the index, just use `.delete()`\nwhich will by default remove the index AND the underlying data.\n\n\n```python\n# Clear all data from Redis associated with the index\nawait index.clear()\n```\n\n\n\n\n    10\n\n\n\n\n```python\n# Butm the index is still in place\nawait index.exists()\n```\n\n\n\n\n    True\n\n\n\n\n```python\n# Remove / delete the index in its entirety\nawait index.delete()\n```\n",
  "tags": [],
  "last_updated": "2026-04-08T12:21:52-07:00"
}