Patronus AI announced it is partnering with MongoDB to bring automated LLM evaluation and testing to enterprise customers. The joint offering will combine Patronus AI’s capabilities with MongoDB’s Atlas Vector Search product.
Today, enterprises are using retrieval-augmented generation (RAG) systems to power key workflows using their internal knowledge base. However, these systems are prone to failure. In past research, Patronus AI found that State-Of-The-Art retrieval systems can frequently hallucinate in real world use cases like financial services. Prior research has also shown that LLMs can struggle with reasoning and numerical calculations.
“Enterprises are excited about the potential of generative AI, but they are concerned about hallucinations and other unexpected LLM behavior,” said Anand Kannappan, CEO and co-founder, Patronus AI. “We are confident our partnership with MongoDB will accelerate enterprise AI adoption.”
Also Read: AI in Online Advertising: Friend or Foe to Traditional Marketing Strategies?
The partnership between Patronus AI and MongoDB brings to market a retrieval system solution that enables reliable document-based LLM workflows. Customers can develop these systems with MongoDB Atlas, and evaluate, test, and monitor them with Patronus AI to increase their accuracy and reliability.
“We recommend developers test iteratively as they experiment with retrieval system design choices,” said Rebecca Qian, CTO and co-founder, Patronus AI. “Patronus AI offers a powerful solution here that is simple to get started with.”
Patronus AI is part of the MongoDB Partner ecosystem.
SOURCE: PRNewsWire