Profile
JarviX: Get through BI+AI! Combining NLP to create a user-centered intelligent decision-making platform Judging from the many No code/Low code platforms on the market, BI (Business Intelligence, Business Intelligence) and AI each occupy two different business fields. The field of BI will rely on a lot of manual data analysis in advance, and then convert the data into a visual infographic, which still requires human input and professional knowledge of front-end personnel. In the field of AI, a large amount of data needs to be input and marked, and the results can be obtained through the combination of various algorithms. However, it is relatively difficult to convert the results into visual charts. 1. What are the main functions of the platform? The main function and advantage of JarviX is that it combines the two together, and is unique in the world. It uses natural language processing (NLP) as a mode of human-computer dialogue, and automatically disassembles the questions asked by users into corresponding programming languages. Use the programming language to obtain data from the database for analysis, and automatically form visual data charts. Huang Yihua believes that the intelligent analysis platform of BI+AI should reach the following four levels: Descriptive analysis: what happened from the past to the present, and describe the current situation. (BI) Diagnostic analysis: including abnormal analysis, which separates normal from abnormal and expresses them in different colors and shapes. (BI) Predictive analysis: According to the trend presented in the data from the past to the present, through the method of machine learning, with the situation unchanged, forecast at the visible time point. (AI) Indicative analysis: If the future trend is known, what should the decision-maker do, provide insights, and enter the modeling stage. (AI) JarviX allows users to express in the most natural way, automatically parses and retrieves data through machine learning, and provides a visual interface. According to the question, it automatically captures and compares the key factors of the database, finds the correlation between each other, assists the user to answer the question in depth from the original relatively vague question, and finally produces a result that can support the questioner. In addition, this series of processes can even be completed in a few seconds, greatly improving efficiency. That being the case, what languages does JarviX understand? In terms of languages, there are currently Chinese, English, Vietnamese, and Cambodian, and Japanese and Thai are also under planning. 2. What preparations should the enterprise make before using it? Need to have data, and know what data the enterprise itself needs, and whether the required information is available, if not, how to obtain it? Huang Yihua mentioned that there are generally three ways to collect data: process software (such as ERP), sensors (temperature, humidity), and machine data. Synergy is different from other No code/Low code platforms in that it has the ability to directly access these three types of data. With JarviX as the center, it intercepts data through the 5G IoT gateway Odin, and the Bifrost IoT platform performs data cleaning. After finishing, it can be put directly into JarviX for analysis. "We need to open up the last mile of data (integration of software and hardware)," Huang Yihua said that if customers do not have data, the test is their own. It is impossible to do data analysis without data, and the answer given without data I can't believe it either. 3. Applicable scenarios and industries? At present, the application of JarviX spans various industries, including electronics, logistics, retail, logistics, metal processing, footwear, semiconductors, semiconductor equipment, banking, life insurance, solar energy, wind energy, energy conservation, and energy storage. In addition, it also depends on the threshold of the industry. For example, banks have very high requirements for information security and financial management, so the threshold is relatively high. They must go through various tests and are only open to a few qualified manufacturers. 4. Is the user limited in capacity? To operate JarviX, you only need to know Chinese input and drag and click, and you can get started immediately after a few hours of course training. The key point is that users need to be very familiar with their own industry domain knowledge to know what data to analyze.
Forum Role: Participant
Topics Started: 0
Replies Created: 0