TL;DR:
- Patsnap, a leader in IP analysis, introduces CoPilot, an AI assistant, to expedite patent and non-patent literature searches.
- CoPilot enhances productivity across Patsnap’s product suite, offering seamless source linking and references through its proprietary LLM.
- Patsnap’s substantial investment and dedicated engineering team underscore its commitment to AI innovation.
- CoPilot empowers IP and R&D teams by providing quick access to vital information, including automatic patent claim summaries and technology insights.
- The AI tool addresses complex queries, translating patents and identifying relevant literature.
- CoPilot also aids in staying updated on dynamic sectors, content analysis, data extraction, and AI security.
- Patsnap’s LLM training incorporates diverse data sources, resulting in exceptional accuracy and in-depth analysis capabilities.
- The company has raised over $350 million in funding, with a broad customer base across various industries.
Main AI News:
Singapore-based IP analysis platform Patsnap, founded in 2007, has consistently revolutionized the field of intellectual property and research and development. Initially renowned for its global patent search database, the company now takes a giant leap forward with the launch of its cutting-edge AI assistant, CoPilot.
CoPilot is set to transform the way IP and R&D professionals operate by significantly accelerating patent and non-patent literature searches. This remarkable innovation promises to streamline IP and R&D workflows throughout Patsnap’s extensive product suite. Leveraging its proprietary LLM (Language Model), CoPilot seamlessly connects to various sources, including journals and patents, while offering valuable references.
Patsnap’s unwavering commitment to advancing AI capabilities is evident in the substantial investments, totaling millions of dollars, and a dedicated team of over 50 engineers. Jeffrey Tiong, CEO and co-founder of Patsnap, asserts that the company’s mission is to eliminate friction from the innovation process for its clients, both within IP and R&D teams and between them. The need for IP analysts and attorneys to conduct prior art and freedom to operate searches is paramount in determining resource allocation. Patsnap’s suite of products is strategically designed to simplify this critical decision-making process.
One notable example is Patsnap’s Analytics product, boasting a vast database of over 180 million patents and 130 million literature pieces from 170 jurisdictions. IP teams can harness Patsnap’s AI tools to gain comprehensive insights into their markets and efficiently safeguard their inventions, according to Tiong.
CoPilot takes Patsnap’s AI offerings to new heights by empowering IP and R&D teams to access essential information more swiftly within patents, non-patent literature, and technical news. Tiong highlights specific benefits, such as automatic summaries of patent claims, technology insights, and links to relevant company patents. Furthermore, CoPilot excels at solving complex queries. For instance, if an R&D professional seeks to enhance battery energy density, CoPilot not only translates patents but also identifies pertinent literature.
In addition to these capabilities, CoPilot offers four more essential features: staying updated on rapidly evolving sectors, conducting content analysis for strategic patents and research, extracting key details from specific patents and literature, and ensuring AI security by not relying on customer data. Tiong emphasizes that CoPilot’s proprietary LLM ensures that customer data remains within Patsnap’s secure firewall, avoiding any exposure to external networks.
Tiong further elaborates on the robust training of Patsnap’s LLM, which draws from a wide range of sources, including patents, academic papers, technical reports, and recent company news, including mergers and acquisitions. Additionally, the model incorporates data curated by IP experts and Patsnap’s own products. This multi-stage learning process enables the LLM to acquire generic, specialized, and aligned data, resulting in exceptional accuracy for both patent and non-patent data. When combined with retrieval-augmented generation techniques, CoPilot’s model surpasses GPT3.5, excelling in in-depth analysis while minimizing the risk of generating incorrect responses.
Patsnap’s impressive journey is reflected in its substantial funding, with investments exceeding $350 million from prominent backers like SoftBank and Tencent. The company boasts a dedicated workforce of over 1,200 employees and a vast clientele spanning various industries, including life sciences, automotive, consumer goods, technology, manufacturing, engineering, and legal. With CoPilot leading the charge, Patsnap continues to shape the future of IP and R&D innovation.
Conclusion:
Patsnap’s CoPilot represents a significant leap in streamlining IP and R&D processes. Its robust AI capabilities, coupled with extensive data sources, position it as a game-changer in the market. Businesses across diverse industries can expect improved efficiency and decision-making in their intellectual property and research endeavors, further accelerating innovation and competitiveness.
