PhnyX Lab launches Korea's first Modular RAG-based solution specialized for medicine and pharmacy
-
2025-01-03
- The first showcase for stakeholders in the pharmaceutical industry took place at Walkerhill Hotels & Resorts in Gwangjin-gu, Seoul.
- PhnyX Lab, a Silicon Valley AI startup, unveiled its Modular RAG-based AI solution to highlight its unique and innovative technology.
- Cheiron, Korea's first generative AI solution tailored for medicine and pharmacy, is anticipated to enhance the productivity of industry professionals.
- CEO Bae Min-seok mentioned, “We will begin with specialized medical and pharmaceutical services, broaden our corporate clientele, and aim to become a leader in the AI market.”
(December 18, 2024) AI startup PhnyX Lab, led by CEO Bae Min-seok, revealed today that it has launched its inaugural industry-specific solution utilizing its unique innovative technology, Modular Retrieval-Augmented Generation (Modular RAG).
The event, held on December 17th at Walkerhill Hotels&Resorts (located in Gwangjin-gu, Seoul), aimed to present Cheiron, a generative AI solution created by PhnyX Lab, to stakeholders in the pharmaceutical industry, such as doctors, pharmacists, and researchers from pharmaceutical companies. The key subjects of the showcase included the formation journey and upcoming strategies of PhnyX Lab, alongside Cheiron's functionalities and development path.
During the showcase, PhnyX Lab's CEO Bae Min-seok presented the company's growth journey and the narrative behind Cheiron, and conducted a demonstration to highlight its technical distinctions from other leading AI solutions. Specifically, Cheiron, grounded in Modular RAG, drew the interest of participants by producing fast and precise responses instantly, in contrast to competing platforms.
█ SK Networks connects with AI talent, fostering unique innovation at PhnyX Lab.
In April of this year, three PhnyX Lab founders, all from Stanford University, including CEO Bae Min-seok, joined forces with SK Networks based on their AI expertise and passion for transformation.
Alongside SK Networks, which possesses a worldwide technology network and capabilities, the founders turned their enthusiasm into a tangible vision by creating AI solutions, validating services, and testing them in the marketplace. Consequently, PhnyX Lab started as a standalone company in September with the aim of transforming the search process through verifiable generative AI solutions and effecting significant shifts in personal work settings and corporate culture.
In this endeavor, SK Networks has harnessed its global AI expertise and connections to assist PhnyX Lab by pinpointing AI and other technology trends both domestically and internationally, as well as through consistent investment and partnership based on Hicosystem, its existing network of specialists from the global technology and investment sectors. Thanks to this support, PhnyX Lab has expedited the progression of Generative AI (Gen AI) solutions and product advancement utilizing Modular RAG, resulting in this demonstration.
Going forward, PhnyX Lab will strive to refine its technology by acquiring clients within the medical and pharmaceutical sectors and improving its international presence through solutions that offer tailored features for individual companies.
█ PhnyX Lab introduces a groundbreaking phase of AI solutions with Modular RAG, its fundamental architectural framework.
Overall, LLM is regarded as a groundbreaking technology compared to current Gen AI since it comprehends the context of inquiries and produces natural responses derived from an extensive dataset. Nonetheless, it has faced criticism for depending on pre-trained data, which may lead to unreliability due to structural constraints, such as misinterpreting information during the learning phase, which results in incorrect answers or offering baseless replies to unfamiliar data.
Under these circumstances, PhnyX Lab has tackled the limitations of LLM by integrating the main benefits of RAG, including real-time searches to minimize learning expenses, enhanced security, and natural language processing technology to select data according to the question's intent. This enables the technology to generate responses based on trustworthy information and to reduce the illusion phenomenon via self-revalidation.
PhnyX Lab went a step further with Modular RAG, enabling internal modules to maintain identical input and output values. Traditional RAG necessitates manual connections between modules to tailor them for various user scenarios, whereas Modular RAG is flexible and easy to connect, allowing for the rapid generation of high-quality responses, thereby facilitating scalability to emerging technologies and fresh data. Similar to Lego blocks, they have a structure that can operate independently or be logically connected depending on the situation. This technology is also useful in improving security since it enables companies to tailor the type and format of data to their requirements, along with optimized storage options for each data type.
By conducting in-depth interviews with experts from the pharmaceutical sector, PhnyX Lab has discovered that its Modular RAG can enhance the article and data discovery process during the research planning phase by as much as 80%. Consequently, Cheiron was created to assist individuals in the pharmaceutical sector in minimizing the time spent searching for documents.
This is the initial industry-focused solution launched by PhnyX Lab, which is anticipated to aid in the democratization of enterprise AI by broadening its services across multiple sectors. PhnyX Lab will continue enhancing its technology while securing clients in the medical and pharmaceutical sector, boosting its worldwide presence with solutions that incorporate tailored features for specific companies.
"Our aim is to establish ourselves as a partner that offers tangible benefits to our customers through solutions that merge expertise and reliability grounded in our core technology, Modular RAG," stated Bae Min-seok, CEO of PhnyX Lab. "Beginning with Cheiron, the cornerstone of PhnyX Lab’s success, we intend to maximize the impact of our technology and enhance business competitiveness for additional companies and communities," he added.
█ "Precise Answers, One-Stop Search" encapsulates the advantages of Cheiron, an AI generative tool focused on pharmaceuticals.
PhnyX Lab’s Cheiron, the first generative AI solution in Korea specializing in pharmaceuticals, presents five unique attributes: expertise in the medical and pharmaceutical fields, comprehensive search capabilities, semantic search functionality, workflow automation, and support for key languages.
As a service focused on the pharmaceutical sector, it understands the keywords of MeSH (Medical Subject Headings), a hierarchical structure for classifying medical scientific data. It employs multiple scholarly search platforms like PubMed and Semantic Scholar and can explore a broad spectrum of resources, including general search engines, other public sources, and internal company files. Cheiron's clients in the pharmaceutical sector can access a variety of databases from various sources via a single platform to gather the results they require.
Furthermore, natural language processing (user-natural language search) technology is utilized to comprehend intent without relying on keyword matching (user-keyword search). Automated processes like data rearrangement and document creation are provided according to search results, allowing pharmaceutical industry employees to enhance productivity and focus on high-value responsibilities, as well as global scalability by accommodating key national languages.
Leveraging these strengths, Cheiron is anticipated to enhance its standing as a generative AI solution tailored for medicine and pharmacy, playing significant roles throughout every phase of the pharmaceutical industry, from research and drug candidate screening to clinical trial planning, production and process development, and market analysis.
A representative from the pharmaceutical sector commented on Cheiron, stating it is a specialized AI offering for medicine and pharmacy with outstanding precision and speed compared to general portals and other LLM services. “If businesses implement PhnyX Lab's Modular RAG system, it will transform into a platform that innovates multiple sectors, such as drug development, surpassing primary retrieval,” they noted.
█ Panel Discussion and VIPs
The event featured a panel discussion titled “Pioneering the Future of Enterprise AI," moderated by Illia Polosukhin, co-founder of NEAR Protocol, a former Google AI engineer, and co-author of the influential paper "Attention is All You Need," which is foundational to ChatGPT, along with being an AI specialist. Accompanying him were Choi Sung-hwan, advisor at PhnyX Lab, and Bae Min-seok, CEO of PhnyX Lab, who discussed the future landscape of enterprise AI, the significance of NEAR Protocol and decentralized technology, among other topics.
Moreover, the Chairman of SK Group, Chey Tae-won, Professor at Stanford Graduate School of Business, Ilya A. Strebulaev, CEO of NOOM and Advisor at PhnyX Lab, Jeong Sae-ju, along with Senior Director at Reddit, Uzair Khan, conveyed congratulatory video messages to endorse PhnyX Lab's groundbreaking work in overcoming the AI bottleneck. The event also saw the participation of doctors, pharmacists, and company researchers—key individuals in the pharmaceutical sector—alongside other leaders from various fields who expressed their genuine interest and congratulations regarding PhnyX Lab's AI innovation.