Kunlun Tech Co., Ltd. has announced the official release of the latest version of its self-developed TianGong AI advanced search feature. This upgrade aims to provide users with more accurate and professional search experiences, particularly in the fields of financial investment and scientific research. Since its launch, TianGong AI has been committed to delivering innovative intelligent search and information processing solutions to global users through leading AI technology.

The four highlights of the TianGong AI advanced search feature include a comprehensive upgrade of multi-level analytical reasoning capabilities, professional AI search for financial investment, professional AI search for scientific research, and intelligent optimization for document AI reading analysis.

WeChat Screenshot_20241105104303.png

These upgrades enable TianGong AI to provide precise search results and transform user needs into executable intelligent answers. The system has been comprehensively upgraded in terms of data and source timeliness, authority, quality, relevance, and content index richness, meeting the needs of users in various fields such as complex problem-solving, financial investment, scientific research, and document analysis.

When handling complex tasks, TianGong AI's reasoning capabilities have made a qualitative leap, providing efficient and accurate answers through a five-in-one approach of problem decomposition, automatic planning, proactive expansion, deep answering, and interwoven text and images. Additionally, TianGong AI boasts exceptional multi-level analytical reasoning capabilities, simulating human thinking and reasoning processes to automatically generate task plans and progressively complete preset task paths.

In the financial sector, TianGong AI's advanced search features include financial policy queries, indicator queries, financial data comparisons, financial statement analysis, company analysis, research report interpretation, and investment management. It can extract key data from complex financial statements, organize them into key indicator tables, support one-click copying, and even compare the relationships between each piece of data and further analyze the reasons.

In the field of scientific research, TianGong AI has established the most comprehensive academic meta-database in China, covering over 200 million English papers, with basic full coverage of academic papers in all disciplines. TianGong AI can directly present precisely matched literature, automatically generate summaries and citations of key information, and support users in quickly obtaining the core content of papers.

In the domain of document AI reading analysis, TianGong AI has made breakthrough progress, not only accurately summarizing and analyzing complex financial statements, research reports, papers, and other documents but also possessing exceptional contextual understanding capabilities, supporting the processing of texts over 500K words long. TianGong AI's PDF document parsing engine and document parsing large model have achieved SOTA in all indicators such as title, author, abstract, citation, images, tables, formulas, and subheadings, especially in multi-column documents and page breaks, where the recognition large model has surpassed all current models.