With the release of ChatGPT, it has become widely recognized that building large-scale language models can perform a variety of tasks, and an increasing number of companies are introducing ChatGPT to improve business efficiency and utilize internal knowledge. On the other hand, some companies may be hesitant to introduce ChatGPT in-house due to concerns about information leaks. This issue may be solved by building your own LLM using an open source LLM that has already been trained in Japanese. In this video, we will compare the advantages and disadvantages of using in-house knowledge with ChatGPT and open source LLM, and explain how to build your own LLM using actual environments and knowledge data. Please take a look.

In the manufacturing industry, the movement to utilize big data generated daily at manufacturing sites in business is accelerating with the aim of increasing sales and improving production efficiency. In particular, "demand forecasting" and "predictive maintenance" achieved by data visualization and analysis and the use of predictive AI, as well as the "introduction of in-house chatbots" using generative AI, will help eliminate human resource shortages, improve customer satisfaction, and reduce costs. It brings many benefits such as reduction. In this video, we introduce examples of "demand forecasting," "predictive maintenance," and "in-house chatbots" put into practice by leading companies in the manufacturing industry that are actually achieving results through the use of data and AI.・We will explain in detail how to start using AI.

Since the 2000s, Apache Hadoop (hereafter, Hadoop) has created a major trend called a "data lake" as a data processing platform for corporate big data. The increasing complexity of data management in terms of perspective has created the challenge of increasing costs. In this video, we introduce a new architecture called "Lakehouse" and its solution as a method to solve the problems of data lakes such as Hadoop. In addition to the advantages of a data lake, such as the handling of diverse data and high-speed search performance, Lake House also has the characteristics of a data warehouse (DWH), realizing "unified data management" and "reduction of operating costs." To do. We also introduce actual examples of data lake house migration from Hadoop, so please use this video as a hint for how to proceed with big data operation and analysis in the future.

[Macnica x Databricks Japan joint session sponsored by Mynavi "DX FRONTLINE ~ Guidepost of Change ~ (held on February 21, 2023)"] In recent years, companies have been using data to visualize "what happened in the past" using BI tools. We are moving from the analysis phase to the phase where AI is used to predict what is likely to happen in the future and utilize it in management decisions. In this session, we will explain the benefits and barriers that the use of both BI and AI brings to business innovation, as well as how to solve them, based on actual advanced cases.