SharePoint to Weaviate

Last updated 22 October 2026

One way: SharePoint stays the system of record, and SPSync keeps a read-only copy of your lists current in Weaviate.

Status: Shipped · product: SPSync for Search & Vector Stores

SPSync replicates SharePoint list items into Weaviate, one object per item, reading from SharePoint and never changing it. The objects carry no vector and each collection is created without a vectorizer, so Weaviate holds them for filters and keyword search until you add an embedding step of your own. SPSync generates embeddings on sync only for Postgres with pgvector and Azure AI Search; doing the same for Weaviate is under consideration, with no plan or date.

Unlike Qdrant and LanceDB, Weaviate ranks keyword matches: every text property SPSync creates is indexed for BM25 by default, so a ranked keyword search over list text works from the first run, with no model involved.

Connecting over REST and gRPC

SPSync needs the REST endpoint, such as http://localhost:8080 or your Weaviate Cloud URL, and the gRPC port, 50051 by default, on the same host. The Python client sends batch imports over gRPC, so open both ports to the sync host. Authentication is an optional API key from an environment variable; a local cluster without one needs none. SPSync keeps a local state file between runs on a stable path.

Collections and properties

A grant applications list, fund_applications, with a multi-choice Themes column gives three collections: fund_applications, fund_applications__themes and the shared users collection. Lookups arrive as an ID and a display value on the object, and a person as a user ID that matches the users collection.

Properties use five Weaviate types: text for text and IDs, int for whole numbers, number for numbers and currency, boolean for Yes/No, and date for dates and times. Weaviate reserves id and vector as property names, so SPSync renames such a column with a _prop suffix. Weaviate refuses to retype an existing property, so SPSync checks first and writes to a new _v2 collection when a SharePoint column changes type. A new column is added as a property in place.

How do I search SharePoint list data in Weaviate without vectors?

SPSync leaves Weaviate’s property defaults alone, so each property is indexed for filtering, and text properties are also indexed for BM25. Text uses Weaviate’s default word tokenisation, which means a filter or keyword query on a text property matches the words in the value. Dates are compared as RFC 3339 timestamps.

A BM25 query with filters, using the Python client: submitted applications that mention community energy, asking for 25,000 or more, changed since April.

import weaviate
from weaviate.classes.query import Filter

client = weaviate.connect_to_local()
apps = client.collections.use("fund_applications")
response = apps.query.bm25(
    query="community energy",
    query_properties=["title", "summary"],
    filters=(
        Filter.by_property("status").equal("Submitted")
        & Filter.by_property("amount_requested").greater_or_equal(25000)
        & Filter.by_property("sp_modified").greater_or_equal("2026-04-01T00:00:00Z")
    ),
    limit=10,
)
for obj in response.objects:
    print(obj.properties["title"], obj.properties["amount_requested"])
client.close()

How do I add vectors to SharePoint objects in Weaviate?

Weaviate 1.31 and later can add a named vector, with a vectorizer and the properties it should embed, to a collection that already exists, including one created without a vector. Weaviate does not vectorise the objects already there when you add it. SPSync rewrites an object each time its row changes and rewrites every object on each reconcile, weekly by default, but it has not been tested against a collection with a named vector added, so try that on a copy of the collection before you depend on it.

The other route keeps the two apart: your own job reads the objects, embeds the properties you pick and writes the vectors to a collection it owns, under the same object IDs, so a hit leads back to the SharePoint item.

Deletes, changes and the scheduled reconcile

A deleted SharePoint item’s object is removed, and so are its objects in the child collections. Each run sends only changed rows; a reconcile, weekly by default, sends everything, which repairs objects changed or removed by hand.

Who reads the collections

Applications already built on Weaviate read list data from the collections without calling Graph, and agent tools use BM25 over them as a keyword search; see AI agents and tools over SharePoint. For retrieval where SPSync produces the embeddings, see RAG on SharePoint lists.

Steps

1. Choose delegated sign-in or per-site Sites.Selected app permissions; the second is what a sync with no signed-in user needs.
2. Set the REST URL, the gRPC port and the API key variable if the cluster needs one.
3. Choose the lists and run; the first run creates the collections.
4. Query them from your application with filters or BM25.

Use a local cluster for development and Weaviate Cloud or a cluster you operate for production. Collection size drives memory and storage, and SharePoint permissions do not travel with the objects, so cluster access is the boundary.

spsync auth login --client-id 00000000-0000-0000-0000-000000000000
spsync export-config --output replication.yaml --pipeline-name weaviate --prefix weaviate --destination-kind weaviate --connection-string-env SPSYNC_DB_CONNECTION --dataset-name weaviate --client-id 00000000-0000-0000-0000-000000000000 --site-url https://contoso.sharepoint.com/sites/example --all-lists
spsync doctor replication.yaml
spsync run replication.yaml

pipeline_name: weaviate
prefix: weaviate
auth:
  mode: delegated
  client_id: 00000000-0000-0000-0000-000000000000
  tenant_id: common
destination:
  kind: weaviate
  connection_string_env: SPSYNC_DB_CONNECTION
  dataset_name: weaviate
scopes:
  - site_url: https://contoso.sharepoint.com/sites/example
    mode: entire_site
reconcile_interval_days: 7
create_reporting_views: false

# every night at 02:00
0 2 * * * cd /opt/spsync && SPSYNC_DB_CONNECTION="$(cat /etc/spsync/db.env)" uv run spsync run replication.yaml >> /var/log/spsync/run.log 2>&1

# .github/workflows/spsync-run.yml
name: SPSync
on:
  schedule:
    - cron: "0 2 * * *"
  workflow_dispatch: {}
jobs:
  run:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: astral-sh/setup-uv@v3
      - run: uv sync
      - run: uv run spsync run replication.yaml
        env:
          SPSYNC_DB_CONNECTION: ${{ secrets.SPSYNC_DB_CONNECTION }}
          SPSYNC_LICENCE_KEY: ${{ secrets.SPSYNC_LICENCE_KEY }}

# azure-pipelines.yml
schedules:
  - cron: "0 2 * * *"
    branches: { include: [main] }
    always: true
steps:
  - script: uv sync
  - script: uv run spsync run replication.yaml
    env:
      SPSYNC_DB_CONNECTION: $(SPSYNC_DB_CONNECTION)
      SPSYNC_LICENCE_KEY: $(SPSYNC_LICENCE_KEY)

# Windows Task Scheduler action (see /docs/cli/automation/)
$env:SPSYNC_DB_CONNECTION = (Get-Content C:\spsync\db.env)
uv run spsync run C:\spsync\replication.yaml
schtasks /Create /SC DAILY /ST 02:00 /TN "SPSync" /TR "powershell -File C:\spsync\run.ps1"

# ~/Library/LaunchAgents/io.spsync.run.plist, then: launchctl load ~/Library/LaunchAgents/io.spsync.run.plist