Create a workspace
Spin up a fabric workspace and load your first dataset.
Start guideGuides, references, and examples to take you from your first query to production agents.
Spin up a fabric workspace and load your first dataset.
Start guideCombine relational filters with native vector search in one SQL statement.
View tutorialWire a model into an agent with memory and decision lineage.
Build an agentSchema design, HTAP, storage/compute scaling, zero-copy clone & time travel.
Ingest and process multimodal data into a queryable knowledge base.
Compose agents, capture decision lineage, run safe A/B tests.
Give agents long-term, governed memory with rollback.
REST and SDK references for every capability, with copy-paste examples in Python, TypeScript, and SQL.
from omnifabric import Client of = Client(api_key="of_live_...") # hybrid search + an agent call, one SDK docs = of.search( query="refund policy", top_k=8, hybrid=True, ) answer = of.agent("support").run( input=question, context=docs, ) # lineage captured automatically
Step-by-step walkthroughs for common use cases.
Reference designs for production deployments.
Engineering deep-dives and product updates.
Join the discussion and get help from the team.
Create a free workspace and run your first hybrid query in minutes.