What You Should Know:
- Artificial intelligence has transformative potential for driving value and insights from data. As we progress toward a world where nearly every application will be AI-driven, developers building those applications will need the right tools for creating experiences from these applications.
- This is why InterSystems is excited to announce the addition of vector search to the IRIS data platform.
InterSystems Announces Integration of Vector Search into IRIS Data Platform
In the realm of large language models, tools such as vector search play a pivotal role in facilitating efficient and precise retrieval of pertinent information from vast datasets. By transforming text and images into high-dimensional vectors, these methodologies enable swift comparisons and searches, even amidst millions of files spanning diverse datasets within an organization.
InterSystems, a leading provider of data management and healthcare solutions, continues its commitment to advancing data processing capabilities closer to the customer’s data without necessitating data transfers to specialized systems. The integration of vector search into the InterSystems IRIS data platform marks a significant milestone in enhancing the platform’s functionality for tasks related to natural language processing (NLP), text, and image analysis. This strategic enhancement enables developers to leverage generative AI for intricate tasks across various applications while ensuring the security of proprietary intelligence.
By empowering the IRIS data platform to manage and query content alongside dense vector embeddings, particularly with RAG (Retrieval Augmented Generation) integration, developers gain agility in adopting new models and use cases. This seamless integration streamlines the development process for generative AI-based applications, ensuring timely responses based on curated data.
One of the prime beneficiaries of this technology is BioStrand, an AI-powered drug discovery company participating in the InterSystems Innovation Program. BioStrand’s Lens AI platform utilizes advanced algorithms, including Large Language Models (LLMs) and patented HYFT Technology, to expedite drug discovery and design processes. HYFTs, acting as unique ‘fingerprints’ in biological sequences, enable precise embedding assignments from different LLMs, forming the basis of a vast knowledge graph encompassing 25 billion relationships across 660 million data objects. This comprehensive graph interconnects biological sequences, structures, functions, and bibliographic information, leveraging cutting-edge technologies like SQL vector search and generative capabilities of LLMs, alongside the semantic expressiveness of knowledge graphs. Through this integration, BioStrand significantly reduces R&D timelines, from development to commercialization, revolutionizing the drug discovery landscape.
