Intelligence Factory has announced the release of OGAR (Ontology-Guided Augmented Retrieval), an AI-driven data retrieval technique that meets the complex data needs of sectors like healthcare, finance, and aerospace. Unlike conventional vector databases and retrieval-augmented generation (RAG) solutions, OGAR uses an ontology-based approach to provide hallucination free, industry-specific insights, tailored for fields like healthcare, finance, and aerospace while keeping sensitive data from being sent to Large Language Model (LLM) providers such as OpenAI, where it might be compromised or used to train new models.
Current AI retrieval tools often rely on an LLM alone, which can lead to “hallucinations” and data leakage. OGAR addresses this with domain-specific ontologies that ground the generative AI, enabling more accurate and contextually relevant data access in specialized fields.
“Many existing AI solutions depend on vector similarity and next token prediction, limiting their effectiveness in complex settings,” said Matt Furnari, CTO of Intelligence Factory. “OGAR grounds the AI with structured meaning, preventing common AI-generated hallucinations.“
Why OGAR Stands Out for AI Agents
OGAR’s ontology-based design is ideal for industries where data accuracy and data security are essential:
- Healthcare: OGAR enables precise access to patient records and health data, without risking HIPAA violations by leaking that data to a third party.
- Sales and Customer Support: For any company that receives or makes phone calls, OGAR can ensure the conversations stay on topic, data is extracted automatically, and can even automate many interactions.
- Aerospace: OGAR streamlines access to technical documentation and legacy databases without exposing that information to the wider internet. It makes lookup of complex relationships and values simple with natural language queries.
