The rapid progression of digital technologies has lent itself to a plethora of policies aimed at bolstering companies that possess core competencies in artificial intelligence (AI) and big data, particularly in their industrial applicationsConsequently, various financial IT service providers have chosen to explore diverse technological innovations in these fields, demonstrating a vibrant atmosphere of creativity and competition.

One such company, Xinyada (stock code: 600571.SH), has made significant strides by focusing heavily on AI integration within the financial sectorWith the advent of a new digitalization cycle, Xinyada's strategy emphasizes foundational upgrades in IT technologyAccording to Ji Jinxiang, the Senior Vice President and Chief Technology Officer of Xinyada, this involves investing in basic technologies to reshape financial institutions' IT systems by deploying state-of-the-art AI technologies, especially those represented by large models.

The core aim is to solidify the underlying technologies essential for AI, particularly in the realm of Optical Character Recognition (OCR) and Natural Language Processing (NLP)—two critical components that facilitate AI models

Advertisements

Launched into the AI field in 2019, Xinyada has made a name for itself by leveraging its expertise in these foundational technologiesDuring the process of big data acquisition in financial institutions, OCR plays a pivotal role, particularly in the various stages of client onboarding and transaction processing.

In practical terms, financial institutions commonly encounter a slew of challenges related to data collection, including the necessity for extensive photo and scan uploads, which can be time-consuming and inefficient if sorted manuallyOftentimes, after the verification process, vast amounts of data lie dormant, with little to no chance of being reusedHence, there’s a growing consensus that, in the age of big models, proprietary data within enterprises has assumed immense valueOCR technology's evolution accomplishes the feat of recognizing increasingly complex documents, thus enhancing data collection and supporting deeper information exploration.

Ji elaborates on this, stating that during the initial collection of information, unstructured data can harbor critical intelligence, such as user intent, behavioral preferences, and market trends

Advertisements

Looking toward the future, these insights are expected to greatly influence operations, client outreach, and risk managementBy combining its superior OCR capabilities with the robust processing strengths of large models' NLP facets, Xinyada intends to transform this data into actionable labels essential in the AI-driven data landscape.

Xinyada's investigative efforts have also been bolstered by its establishment of the Financial Big Data and Artificial Intelligence Research Institute in 2016. Since 2022, the company's strategic focus has been increasingly centered on financial contextualization, digital innovation, and the implementation of intelligent systemsThis institute has paved the way for extensive research into various capacities, including natural language processing, image recognition, knowledge graphs, and customized AI modeling.

After the extensive collection and tagging of data, another emphasis in the digitalization process is the synthesis and output of this information through AI, with NLP technology being a vital driver in this transformation

Advertisements

Early adopters of NLP technology in the financial services sector developed applications such as chatbots and AI assistantsNow, with the emergence of large language models, pre-trained models can generate outputs that are more coherent and fluent than ever beforeDevelopments in generative AI, exemplified by models like ChatGPT, have elevated the technology as a significant vector of productivity; however, the discussions surrounding its safety and efficacy remain contentious within industry circles.

Concerns surrounding the reliability of AI outputs were raised by Zhou Xiang, a distinguished researcher at Zhejiang University's Guanghua Law School, who highlighted the risks tied to the quality of training data utilized by AI systemsIn domains like healthcare, finance, transportation, or law, faulty outputs could lead to serious ramificationsIn response, Xinyada remains cognizant of these challenges while developing generative AI applications

• China-U.S. Economic Cycle Repositioning
• Profitability Starts with Loss Prevention
• Storage Market Recovers: HBM Gains Popularity
• AI Integration Reshapes Financial IT Services
• Power Battery Recycling Industry Set for Billion-Dollar Market

Ji underscores that the financial sector presents two crucial hurdles: First, large models originate in consumer-oriented internet environments, yet financial entities demand stringent privacy protection and information security protocols, necessitating localization and controllable solutionsSecondly, due to the inherent phenomenon known as "hallucination," the accuracy of output in AI-driven interactions holds significant implications, especially considering service-level liabilities for financial institutions.

To ensure reliability and integration into existing operational frameworks, extensive remapping and fortification of the large model technologies are paramountThe synergy of technology and business realization is emphasized as the key to unlocking their true potentialThere exists an emerging competitive landscape among financial IT services pursuing advancements in AI and big data capabilities—where success hinges upon the effective amalgamation of underlying technologies with systemic solutions.

Since 2018, industries have rallied around the national government's three-step strategy for localization, striving to transition towards self-driven technologies

The inception of the first pilot program for the financial sector in August 2020 involved 47 participating organizations, comprised of banks, insurers, and brokerage firms, mandating that these entities allocate 5-8% of their outsourced IT expenditures towards domestic software and hardware solutionsBy May 2021, the pilot program expanded to 198 organizations, transitioning to a fully self-reliant framework for office automation and email systems and requiring them to commit 15% of yearly IT expenditure to the new standardized productsIt’s evident that self-reliant technologies are gaining momentum throughout the finance sector.

As Xinyada amplifies its focus on research and development, the pursuit of integrating AI into its operations continuesThe 2022 launch of Xinyada's AI strategy is predicated on reconstruction of financial operations and management processes utilizing next-generation technologies like AI and big data, all aimed at lowering barriers to innovative technology for clients while empowering broader financial ecosystems.

Ji notes that the embedding of foundational AI technologies has reshaped product offerings, enabling an iterative upgrade of traditional solutions that incorporate AI capabilities