Finch AI, developer of AI-based solutions that deliver rich, entity intelligence in real-time, announced that it has built and deployed two new agentic AI solutions designed to enhance its ability to quickly add depth and dimension to text analytics missions.
“We’re so pleased to be able to use these agents to expand our entity knowledgebase and, by extension, the accuracy and robustness of the results we return to users,” Finch AI Chief Technology Officer Scott Lightner said. “These two agents, together, are a powerful way to ensure our solutions are even more responsive to users and the entities that matter to them and their missions.”
Also Read: Polestar Analytics Secures $12.5M to Boost AI & Data Platform
Available now in Finch for Text® and Finch Analyst® the two new agents are:
- Entity Creation Agent: This agent is set in motion when an entity appears in text and is extracted as an entity, but we are unable to resolve it to an entity in our knowledgebase. For example, when the name of a newly opened business appears in a piece of text, Finch AI extracts that name and uses context to determine that it is a business, but given that it is a recently formed entity, there would be little to no information of it in our knowledgebase. Today, when this occurs, the Entity Creation Agent automatically conducts detailed research to find information about the new business. It draws on Finch AI’s agent toolbox, data validation agents, and a large language model (LLM) to create a summary of the entity, including images from the web. The agent then creates a knowledge base entry inclusive of the information it finds and suggests other potentially related entities.
- Relationship Vetting Agent: Finch AI’s ability to accurately and dynamically discover relationships between entities is something customers depend on. We offer 90+% accuracy and have more than 180 million relationships in our knowledgebase. When a new relationship between two entities is discovered – for example, subsidiary-of or employer-of – the Relationship Vetting Agent verifies the validity of that discovery. It uses agent tools to perform research, Finch knowledgebases and an LLM and returns a justification to users as to why the relationship is correct, or not. This justification describes the nature of the relationship and furthers the users’ ability to trust and rely on Finch AI’s relationship outputs.
“At a time when many AI companies are just beginning their use of agentic AI, these two agents enable Finch AI customers to tap into the enormous potential of agentic AI in a truly meaningful way,” Lightner continued. “Moreover, they are powerful agents for generating trusted data for knowledge graphs and represent the first of many agentic AI innovations customers can expect from us.”
Source: PRNewswire
