Onthos
Onthos builds a living ontology for financial markets, transforming fragmented information from a firm's sources into connected, evidence-linked intelligence. The platform distills raw files into typed facts linked to their original passages, enabling users, applications, and AI agents to query across the entire corpus with higher accuracy than RAG at a fraction of the LLM compute cost. It is trusted by index providers, banks, asset managers, and hedge funds.
- Data & Analytics
- Financial Technology
- Artificial Intelligence
- Software Only
Funding
Funding not disclosed
Founders
Founder details are not available yet.
Product
Problem
Financial firms often have fragmented information across markets and within the firm, leading to trapped knowledge, partial answers, and repeated costs. Investment teams need to answer questions that span multiple sources, but raw files and unstructured data make it difficult to extract and connect facts efficiently.
Solution
Onthos builds a living ontology from a firm's information sources, turning them into connected, evidence-linked intelligence. The platform processes raw files, extracts typed facts, and links each fact to its original source, creating a knowledge graph that evolves as new information arrives. Users, applications, and AI agents can then query across the entire corpus, retrieving linked facts and evidence without rereading raw text for every query. This approach reduces the cost of information retrieval while enabling more accurate, context-aware answers, with significantly higher accuracy than RAG at a fraction of the LLM compute cost of reading documents directly.
Target Audience
Primary customers are index providers, banks, asset managers, and hedge funds that need to connect fragmented market and internal data for investment research and decision-making.
Features
- Builds a living ontology from a firm's sources, including companies, products, technologies, business models, and themes, with relationships and metrics
- Distills information into typed facts, each linked to the original source passage for traceability
- Supports querying by users, applications, and AI agents across the whole corpus, with evidence accumulation
- Includes an NLP pipeline designed around the ontology to process raw files and serve facts at scale
- Uses the Theia Insights Industry Classification (TIIC) taxonomy for industry and theme classification
- Provides a distilled summary view and typed facts with hover/tap functionality to bring source documents to the front