neocarta vs slayer
Neo4j Labs' semantic layer for data agents: ingest warehouse schema, business glossary, metrics and query history into one graph, then serve it over MCP so agents route queries and write grounded SQL. — versus — Embeddable semantic layer for AI agents: define metrics once, compose them with expressions and time shifts, row-level security and read-only SQL, over MCP, REST, CLI, Python or a Postgres facade.
Both are semantic layers served to data agents over MCP; neocarta models schema, glossary and query history as a Neo4j graph for routing, SLayer compiles metric expressions into SQL with row-level security.
| neocarta | slayer | |
|---|---|---|
| Stars | 140 | 227 |
| Forks | 26 | 31 |
| Language | Python | Python |
| License | Apache-2.0 | MIT |
| Last activity | 7 days ago | today |
| Topics | knowledge-graphs, data | data |
| Curated connections | 8 | 5 |
neocarta — the curator's take
The honest fix for Text2SQL: the model is not bad at SQL, it is blind to your data landscape. Neocarta pulls schema metadata, foreign keys, sample values, glossary terms, governed metric definitions and real query history into a Neo4j graph — only metadata crosses over, data stays in the source — then serves it to agents over MCP with full-text, vector and hybrid search that returns columns, types, example values and the FK references needed to build a join. Embeddings are optional; catalog search works from schema alone. It ships a runnable LangGraph + BigQuery agent so you can see the routing loop end to end. Read the label though: Neo4j Labs, explicitly experimental, not a supported product, 98 stars, and it assumes you are willing to stand up and maintain a Neo4j instance next to your warehouse.
slayer — the curator's take
Reach for SLayer when agents keep writing plausible-but-wrong SQL against your warehouse: define columns and metrics once, and agents compose them (ratios, time shifts, alternate aggregations, multi-stage queries) through a search, inspect, query flow instead of free-hand joins, with read-only connections and row-level security handled for them. It imports dbt and Cube definitions, embeds as a Python library, and speaks MCP, REST, Flight SQL and the Postgres wire protocol, so BI tools hit the same definitions. It is the core of Motley's product and still young (a few hundred stars, six contributors). Skip it if you want dashboards out of the box (wrenai) or a layer that builds itself from your existing BI assets (ktx).