Architecture

Amazon Neptune – A fully managed graph database service that makes it easy to build and run applications that work with highly connected datasets. An Amazon Neptune-based knowledge graph finds relationships between projects and provides recommendations.

Amazon Bedrock – A fully managed service providing API access to high-performing foundation models (FMs) from Amazon and leading AI companies. It supports fine-tuning, RAG-based customization, and Automated Reasoning — without managing infrastructure — with built-in security, privacy, and responsible AI capabilities for building and scaling generative AI applications on AWS.

Amazon Bedrock AgentCore – A managed platform for building, deploying, and operating enterprise-grade AI agents at scale. It provides purpose-built infrastructure, session management, memory, and security controls to move agents from proof-of-concept to production — enabling secure, reliable autonomous agent operations at enterprise scale.

Ontop – An open-source Virtual Knowledge Graph (VKG) system that exposes relational databases as virtual RDF knowledge graphs — without moving data. It translates SPARQL queries over ontology-defined knowledge graphs into SQL executed directly against relational sources, using R2RML mappings and lightweight OWL ontologies. Built on OBDA (Ontology-Based Data Access) principles, Ontop enables semantic querying over existing data infrastructure with no data duplication.

Amazon OpenSearch Service – A managed service for real-time search, monitoring, and analysis. It provides vector storage and semantic similarity search for embedding-based retrieval in RAG and GraphRAG workflows. Available as provisioned clusters or OpenSearch Serverless (auto-scaling, pay-per-use).

Interface Layer

Intelligence Layer

Knowledge Layer

Data Layer

Orchestration Layer

Governance / Security