Skip to content
Back to projects

Project

OrgMind

Centralized agentic data layer using hierarchical knowledge graphs for cross-system orchestration.

Neo4jKnowledge GraphsGraph RAGFastAPI

Overview

OrgMind is a centralized data layer that unifies information across an organization's systems into a hierarchical knowledge graph. It enables AI agents to query, reason over, and act on cross-system data through a Graph RAG architecture.

My Role

Sole architect — designed and built the entire system from scratch, including the graph schema, ingestion pipelines, RAG layer, and API surface.

Approach

  • Designed a hierarchical knowledge graph schema in Neo4j that maps organizational entities, relationships, and workflows.
  • Built data ingestion pipelines that normalize and connect data from disparate enterprise systems.
  • Implemented Graph RAG — combining graph traversal with retrieval-augmented generation for contextually rich agent queries.
  • Exposed a FastAPI service layer for agent orchestration, enabling other systems to query the knowledge graph programmatically.

Outcome

Deployed as the central data backbone for agentic workflows, enabling cross-system orchestration that was previously impossible with siloed data stores.