Mecletter May｜We will tell you how to raise your child (AI) well. First, cloud and data…
Date of creation : 2024-05-30
Hello, dear readers of the Meckletter!
It's the last week of May, and it's hard to let go of the blessings of the holiday season.🙇
This month, the largest IT conference in the country, AWS Summit Seoul, was held! I believe many of you participated as well, and there were plenty of discussions related to 'generative AI' from the session halls to the expo.
At the Megazone Cloud booth, we visualized the steps from cloud migration, which is essential for successful generative AI implementation, to AI development as the ‘AX lifecycle’. Given the high level of interest at the expo, I would like to delve into this in detail in this Meckletter for our curious subscribers.😉
Generative AI, is the valley of disappointment coming? 😨
(Source: Gartner, What’s New in the 2023 Gartner Hype Cycle for Emerging Technologies)
Before we take a closer look at the ‘AX lifecycle’, I want to show you something! Are you familiar with Gartner's hype cycle? The hype cycle is used as an indicator of the maturity and market adoption of new technologies, consisting of five stages: technology trigger → peak of inflated expectations → trough of disillusionment → slope of enlightenment → plateau of productivity.
In the report ‘2023 Hype Cycle for Emerging Technologies’ released last year, Gartner classified ‘generative AI’ as being at the peak of inflated expectations among the five stages of the hype cycle. Interestingly, they predicted that next year, it would enter the ‘trough of disillusionment’ stage, where market expectations and interest sharply decline according to the cycle.
Why do new technologies fall from the peak of expectations into the trough of disillusionment? This happens because companies that adopted the new technology early on fail to monetize it, leaving only a few to continue investing. The logic is that after good revenue cases emerge and second to third generation products are released, they go through the slope of enlightenment and eventually enter the plateau of productivity where they are widely adopted in the market.
Companies attempting this will all want to successfully adopt new technologies in their early stages and enjoy explosive value from innovation. So, what is needed for successful AI adoption that survives even when the trough of disillusionment comes?
A journey of a thousand miles begins with a single step: from cloud migration to operation
STEP1️⃣ Cloud Migration
You may have heard many times that cloud migration is essential for generative AI adoption, and this is due to the ‘flexibility’ and ‘accessibility’ that cloud environments provide. To develop and establish a customized model that aligns with business goals, it is necessary to train the model with each company's data, which requires vast amounts of data and computing power. The cloud is suitable for these requirements as it allows resources to be easily scaled up or down as needed.
According to the <2024 Domestic Cloud Computing and AI Status and Outlook> published in April this year, 75.4% of domestic companies are utilizing the cloud in some way. However, upon closer examination, the most common response was that ‘only some non-mission-critical tasks are run in the cloud’ at 36.6%, while only 5.9% reported that ‘all tasks are run in a cloud environment’.
While the cloud has operated as a massive technological trend over the past decade, and its necessity has become common knowledge, the journey must accelerate beyond digital transformation (DX) to AI transformation (AX) in this era.
In large-scale migration, it is crucial to clearly identify the organic relationships between applications and databases, as well as areas that require modernization, such as OSS transitions. The above image presents the Mass Migration & DR Framework (hereinafter referred to as MMF) proposed by Megazone Cloud's Mass Migration & DR Center (hereinafter referred to as MMDR Center), which consists of a total of 8 service frameworks that must be utilized to design the cloud journey from Discovery to DX.
The MMDR Center of Megazone Cloud has developed a cloud migration solution called Hyper Mig based on the know-how accumulated through years of cloud transfer experience. The process of cloud migration is broadly divided into three stages: Discovery (inventory collection and validation) – Assessment (integrated analysis of infrastructure, applications, and databases) – Migration. Hyper Mig is an automated tool that efficiently collects a company's inventory and deeply diagnoses and analyzes the relationships between infrastructure, applications, and databases. Through this, companies can gain visibility into their cloud transition projects and shorten the duration.
👉Cloud Migration Solution, Hyper Mig
STEP2️⃣Cloud Operations and Management
Today, as cloud technology becomes more widespread, more companies are focusing on cloud operations and management. As most companies adopt multi-cloud environments operating multiple CSPs, they face various challenges such as excessive cloud spending, increasing cloud security threats, managing multiple cloud accounts, and scattered cloud assets, making efficient management another critical task.
Megazone Cloud offers a multi-cloud integrated portal, SpaceONE, which allows for the unified management of various cloud resources such as costs, failure alerts, and security from a single platform. Beyond managing cloud resources and costs, it enables systematic management of organizations and projects using workspaces.
👉Multi-Cloud Integrated Portal SpaceONE
STEP3️⃣Cloud Security
Do you remember the discussion about cloud security in the era of generative AI we had in the April Mekletter? The attack surface is increasing day by day due to threats from hackers and user threats caused by 'shadow IT.' Therefore, what is needed is to internalize the security capability to analyze and detect large amounts of logs and event data in real-time based on AI.
However, actively detecting such threat events and analyzing threats and vulnerabilities is not easy. Receiving remote security monitoring services can also be a burden for companies, and it is challenging to have the expertise required for security monitoring.
The Cloud Technology Center of Megazone Cloud supports the establishment of a security dashboard environment that detects log events occurring in security services within cloud workloads, allowing security personnel of each company to easily analyze them.
👉Consult with Cloud Security Experts
AI is Better Together: From Data Analysis to Development
STEP4️⃣ Data Analysis
When discussing AI projects, one cannot overemphasize the importance of 'data.' High-quality data collected over a sufficient period is essential for the successful implementation of AI.
However, as the volume of data within companies surges and systems become more distributed, effective integration, management, and utilization of data are becoming increasingly difficult. It often requires a significant amount of time and effort to request and verify the necessary data.
Based on this need, Megazone Cloud is launching the AI & Data Analysis platform 'DP360'. 🎉 From data exploration to AI analysis and visualization, it creates a self-service environment that enables users to carry out the entire data analysis process without support from IT personnel.
It consists of the catalog solution 'DP360 Catalog,' which allows users to freely explore high-quality data through unified data supply and sharing while controlling access based on permissions to provide data governance functions, and the 'DP360 Portal,' which offers a self-service environment by automating data supply, analysis, model serving, and visualization service integration from preparing the analysis environment.
The foundation of generative AI that provides accurate answers is high-quality data and data governance. By adopting DP360, high-quality data can be supplied to generative AI to improve the quality of results, and access permissions for data can also be applied to enhance the security of sensitive data.
👉AI & Data Analysis Platform, DP360
STEP5️⃣ AI Development and MLOps
If the cloud infrastructure is well-established, the data is well-refined, and specific goals for the implementation of generative AI have been set, it's time to move into the practical stage of developing and embedding AI! However, many challenges still exist here. In particular, the lack of skilled human resources for building and operating AI applications or high development costs often act as barriers.
(Source: nvidia blog)
At this point, the role of MLOps is crucial as it automates AI workloads, allowing developers to easily handle everything from AI development to management. MLOps is a combination of ML (Machine Learning) + Ops (Operation), aiming to integrate development (Dev) and operations (Ops) to enhance both development productivity and operational stability, similar to DevOps, by integrating ML models into the software development process to enable continuous deployment and maintenance.
Megazone Cloud offers the Matilda AI Platform, which streamlines and automates AI workloads, allowing teams to focus on practical tasks in the field. Through the Matilda AI Platform, the entire Machine Learning Lifecycle can be managed from model development, automating the training (CT) and deployment (CD) processes of ML models.
AI engineers can use the development environment without the cumbersome procedures involving AI-specific processors like GPUs and IPUs, and MLOps pipelines can automate essential tasks such as training, hyperparameter tuning, and inference deployment, enabling continuous and repetitive execution.
👉Optimal AI Integration Platform, Matilda AI Platform
🌎Go Global🌎
Are you interested in expanding overseas?
Let's read about the trends in the global IT market too!
[Japan] Meet promising startups from Japan at NextRise2024!
A festival of startups! NextRise 2024🎪
As part of the GTM business, we are inviting promising startups from Japan in collaboration with Megazone Japan and AWS Japan. If you would like to meet with Japanese startups, please contact us at gojapan@megazone.com. For details on booth consultation scheduling and Korean-Japanese interpretation, please check the [main text].
[Japan] Korea Tourism Organization × WeWork KOREAN STARTUP NIGHT 2024: Will AI replace human workers in Vietnam?
In April, we participated in the KOREAN STARTUP NIGHT held at WeWork Shinjuku Link Square by Megazone Japan, and you can check out the atmosphere, the contents, and our impressions [main text].
[Southeast Asia] Southeast Asia's data center market surpasses $14 billion!?💰
It is said that by 2028, the Southeast Asian data center market will reach $14 billion. If you are curious about why global companies are moving to Southeast Asian data centers, please check the [main text] for detailed information.
[Vietnam] Security camera incidents💣Vietnam government's response measures💊
According to statistics from Pavana (a Vietnamese camera manufacturer) in October 2022, 90% of the 5 million surveillance cameras imported annually into Vietnam are Chinese brands. If you are curious about how the Vietnamese government responded to security camera incidents, please check the [main text].
👉For inquiries about global expansion, contact us here!
Vietnam: ask_vietnam@megazone.com
Japan: gojapan@megazone.com
