One engine. Three workloads.

Full-text search, knowledge graphs, and columnar analytics. Unified behind standard SQL, designed to embed inside your product.

The Engine

Query everything with standard SQL

One interface for search, graph traversal, and analytics. No separate systems to integrate.

Browse and manage datasets through the application or connect directly via SQL. Full-text search with semantic and keyword matching, graph queries across entity relationships, and columnar analytics across millions of rows.

Example: Legal teams join internal contracts with public caselaw for precedent research. Cybersecurity teams correlate threat data across logs and entity graphs. All with standard SQL.

See the technical details

Dataset list view with tags and descriptions

Browse and search your datasets

AI Orchestration

Run AI workloads from SQL

LLMs, embeddings, classification, and code execution. Callable as functions, no external orchestration layer.

Define pipelines that run language models, generate embeddings, classify text, or execute Python. Call them as SQL functions to build materialized views augmented with AI logic. Configure inputs, monitor runs, and iterate on results.

Example: Run classify(body, 'risk', 'compliance', 'standard') inline, generate embeddings with embed(body), or summarize documents with summarize(body, 100). All from SQL.

See the technical details

Pipeline list view with status indicators

Manage all your pipelines

Built on open standards

No vendor lock-in. No proprietary query languages. Export anytime.

Open formats

Parquet and Arrow. Your data stays portable. No proprietary storage formats.

Standard SQL

ANSI SQL:2011 with PostgreSQL-compatible syntax. Every engineer already knows it.

Cloud or on-prem

Hosted by us, or deployed on your infrastructure. Compute and data stay where you need them.

See the engine in action

Book a technical demo or start building with the free developer tier.

Book a Demo Read the Docs