Repolode
Submit
semantica-agi

semantica-agi/semantica

Python infrastructure that turns enterprise data into a graph-based context layer with ontologies, reasoning, provenance and decision audit trails for AI systems.

View on GitHubWebsite
Stars
14,133
Forks
1,728
Open issues
131
Last commit
yesterday

What is it?

Semantica is a Python library for building knowledge graphs and Context Graphs that sit beneath an LLM, vector store, or agent framework. It ingests data from files, databases, and enterprise platforms such as Databricks, Snowflake, and SAP, extracts entities, relations and events, resolves conflicts and duplicates, and stores the result as a queryable graph. Ontologies (OWL, SHACL, SKOS) define what entities mean, not just how they are embedded.

The project focuses on explainability at the system level. Decisions are stored as first-class graph nodes with causal links, precedent search, and impact analysis. Provenance is tracked with W3C PROV-O, and results can be exported to RDF, OWL, JSON, CSV, Parquet or Cypher. Reasoning uses forward chaining, a Rete network, Datalog and SPARQL, and the core pipeline is deterministic, so an LLM is optional rather than required.

Storage is pluggable across RDF stores (Oxigraph, Jena, RDF4J) and property graphs (Neo4j, FalkorDB, Neptune), with several vector store backends. It also provides a CLI, REST API, MCP server, and integrations with Agno, CrewAI and LangChain.

Key features

  • 01Builds knowledge graphs from files, databases, and enterprise data platforms
  • 02Extracts named entities, relations, events and triplets from text
  • 03Records AI decisions as graph nodes with causal chains and precedent search
  • 04Tracks provenance with W3C PROV-O and exports to RDF, OWL, JSON and CSV
  • 05Runs deterministic reasoning with forward chaining, Rete, Datalog and SPARQL
  • 06Supports swappable RDF and labeled property graph backends

Use cases

  • Building a governed knowledge graph from data already in a lakehouse or warehouse
  • Adding auditable decision records and explanations to AI agents in regulated industries
  • Resolving duplicate entities and conflicting facts across multiple data sources
  • Serving as a self-hosted graph layer for GraphRAG pipelines

Quick start

  1. 1

    Install the package from PyPI

    pip install semantica
  2. 2

    Verify the install

    semantica doctor

Things to know

  • Documentation and tooling are broad; the README covers many modules, so the core scope may take time to learn.
  • Many advanced features (external graph backends, LLM providers, enterprise connectors) need extra setup beyond the base install.

Topics

#knowledge-graph#llm#ontology#graph-rag#provenance#reasoning#agent-memory#python#ai#ai-governance

FAQ

What is semantica?
Python infrastructure that turns enterprise data into a graph-based context layer with ontologies, reasoning, provenance and decision audit trails for AI systems. Semantica is a Python library for building knowledge graphs and Context Graphs that sit beneath an LLM, vector store, or agent framework. It ingests data from files, databases, and enterprise platforms such as Databricks, Snowflake, and SAP, extracts entities, relations and events, resolves conflicts and duplicates, and stores the result as a queryable graph. Ontologies (OWL, SHACL, SKOS) define what entities mean, not just how they are embedded.
Is semantica free to use? What license does it have?
semantica is open source under the MIT license. Check the license terms before using it in commercial products.
Is semantica still maintained?
Actively maintained. The last commit to semantica was on Oct 19, 2026.
What can I use semantica for?
Building a governed knowledge graph from data already in a lakehouse or warehouse Adding auditable decision records and explanations to AI agents in regulated industries Resolving duplicate entities and conflicting facts across multiple data sources Serving as a self-hosted graph layer for GraphRAG pipelines
How do I install semantica?
Install the package from PyPI: pip install semantica Verify the install: semantica doctor
Where can I get semantica?
The source code and installation instructions are on GitHub: https://github.com/semantica-agi/semantica

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects