After taking a picture with your smartphone, if you search for “dog,” pictures of your dog will be displayed in a row. Behind this experience, which has become commonplace, there is actually very advanced technology at work. The photo doesn't have a “dog” tag. AI analyzes the image and determines “this is a picture of a dog.”
ApertureDB promoted by Google realizes this mechanism at an enterprise level and at a much higher level. This database, which began to be offered on the Google Cloud Marketplace in 2024/10, overturns conventional common sense of data management and is about to establish a new standard for the AI era.
What exactly is ApertureDB? Why is it getting so much attention now? Let's explore its innovation and possibilities with familiar examples.
1. What is ApertureDB? Let's understand it with familiar examples
The best way to understand ApertureDB is to start with our everyday lives.
When you open Netflix, it'll recommend movies and TV shows that match your preferences. When you look at a product on Amazon, “People who have seen this product have also seen this product” and related products are displayed. On Instagram and TikTok, posts that you might be interested in appear one after another.
What these services have in common is that they combine and analyze various types of data, such as text, images, video, voice, and user behavior history. Conventional databases were good at structured data such as numbers and letters. However, what is required in the AI era is to efficiently process “multimodal data” ** in different formats, such as photos, videos, audio, documents, etc., in one system.
ApertureDB was created to solve exactly this problem. It is a database system that integrates AI technologies such as image recognition, natural language processing, and speech analysis to understand relationships between data in different formats and can be searched at high speed.
For example, hospitals can manage radiographs, patient symptom records, past diagnosis history, doctors' notes, etc. with a single system, and “past treatment cases for patients with similar symptoms” can be instantaneously searched. Data previously managed by separate systems can be integrated by ApertureDB to provide strong support for doctors' diagnoses.
2. Why is ApertureDB in the spotlight now
Explosive data growth is plaguing enterprise IT departments. According to a survey by IDC (International Data Corporation), it is predicted that the amount of data in the world will reach 175 zettabytes (1 zettabyte = 1 billion terabytes) by 2025. Moreover, over 80% of it is unstructured data such as images, video, audio, and documents.
Let's think of a traditional enterprise system. The customer database is neatly arranged with names, addresses, and purchase history. Product names, prices, and inventory numbers are recorded in the product database. However, product photos posted by customers on SNS, voice inquiries sent to call centers, product review videos, etc. are often managed by separate systems or are not used in the first place.
Huge business opportunities lie dormant here. For example, consider a company that operates a fashion shopping site. A customer posted “these clothes are cute!” on Instagram If a post with a photo called, the customer's past purchase history, and stock status of similar products can be analyzed by combining them, the accuracy of personalized product recommendations will improve dramatically.
However, in reality, the technical hurdles were too high. In order to link image recognition systems, natural language processing systems, and conventional database systems, specialized knowledge and enormous development time were required. According to the ApertureData survey, it is said that it took an average of 6 to 9 months for companies to build such multimodal AI systems.
ApertureDB solves this problem from the ground up. It is possible to build an environment where data in different formats can be managed in a single system, automatically learns the relationships between them, and can be searched at high speed in just a few minutes.
3. The three core functions of ApertureDB
ApertureDB's innovation lies in the integration of three core features. Let's try to understand these with familiar examples.
3-1. Multimodal data management: “organizing” the digital world
Imagine organizing your home photos. In the old days, paper photos were attached to albums and handwritten notes were attached. In the digital age, photos are now stored separately on smartphones, videos in the cloud, and memo apps.
ApertureDB's multimodal data management is a function that aggregates this scattered digital information in one place and organizes it while maintaining relationships with each other. Data in different formats, such as text, images, video, audio, and documents, can be managed within the same system.
The key is not just storing, but automatically understanding relationships between data. For example, product photos, product descriptions, customer review videos, purchase history, etc. are automatically grouped as “information related to this product.”
3-2. Vector search: technology to instantly find “similar”
If you ask Spotify to “find a song similar to this song,” they will suggest songs with similar melodies and rhythms. This is a familiar example of vector search.
Conventional database searches were based** “exact match”. If you search for “Tanaka Taro,” you will only find “Tanaka Taro.” But in the AI world, it's important to find “similar” ** things.
ApertureDB's vector search quantifies the** “meaning” and “characteristics” of the data** and calculates similarity. For example, if you search for an image of a “red sports car,” you can find images of cars with similar colors and shapes regardless of manufacturer or model. What's more, it can process 2 to 4 times faster** than traditional systems.
3-3. Knowledge Graph: Visualize the “connections” of information
“Maybe they know each other?” on Facebook or LinkedIn Think of someone who is suggested. This is the result of analyzing the**"connections"** between you and your mutual friends, work, school, etc.
ApertureDB knowledge graphs apply this** “connection” concept to business data. Using elements such as customers, products, transactions, regions, and time as points (nodes), we create a network structure** connecting these relationships with lines (edges).
For example, it will be possible to search with complex conditions such as “products that women in their 30s living in Tokyo tend to buy in spring and have become a hot topic on SNS.” Until now, tasks that were analyzed by hand across multiple systems can now be executed with a single query.
4. The power of ApertureDB seen from actual use cases
ApertureDB has already been put into practical use by many companies, and its effects are showing in numbers. Let's take a look at its power through specific examples.
4-1. Retail Industry: Addressing Badger Technologies' “Product Placement Mistakes”
Have you ever seen PET bottles of tea sneaking into cola shelves in supermarkets? This is a major challenge in the retail industry called a** “product misplacement” **.
Retail automation company Badger Technologies tackled this issue by introducing ApertureDB. Images of product shelves taken by robots patrolling the store, product database information, past placement error history, etc. are integrated and analyzed.
The results were astounding. The performance of vector similarity search has been improved by 2.5 times, and the accuracy of detecting placement mistakes has been greatly improved. Tasks that previously relied on human eyes can now be monitored 24 hours a day, 365 days a year by AI.
4-2. Video analysis industry: Zippin efficiency examples
Zippin, which provides video analysis technology, needed to train an AI model with hundreds of thousands of labeled videos. In conventional systems, video files, label data, and metadata are managed separately, and it took a huge amount of time just to prepare the data.
With the introduction of ApertureDB, Mr. Hareesh Kolluru, the company's director in charge of AI/ML, testifies that “we were able to reach the goal twice as fast with half the resources.” The integrated management of videos, labels, and metadata has dramatically improved the efficiency of model training.
4-3. A revolution in quality control in manufacturing
An auto parts manufacturer uses ApertureDB for visual inspection of products. Early detection of defective products and quality improvement are realized by integrating and analyzing product photos, inspector comments, past defective product data, repair history, etc.
In the past, quality judgment, which relied on the experience and intuition of skilled inspectors, has been standardized by AI, and high-precision inspections are now possible even for newcomers. Furthermore, through trend analysis of defective products, improvements in the manufacturing process have also been clarified.
What these cases have in common is that new value is created by integrating diverse data that was previously managed separately. ApertureDB is more than just a data storage system; it functions as a strategic tool that enhances business competitiveness.
5. Why Google recommends ApertureDB
The news that ApertureDB was launched on the Google Cloud Marketplace on 2024/10/28 caused major ripples in the IT industry. Why does Google actively promote databases developed by external companies on its own cloud platform?
5-1. A new battleground for AI competition
The current AI competition is shifting from simply competing for model performance to a stage where implementation capabilities such as “how practical AI systems can be constructed” are contested. Large-scale language models like ChatGPT and Gemini are becoming less of a technical differentiator.
True differentiation depends on whether an AI system can be built to solve actual business problems using the company's actual data. And to that end, a high-quality data management platform is essential.
For Google's cloud business, ApertureDB is a strategic partner to accelerate AI adoption by client companies. Dai Vu, the person in charge of Google Cloud Marketplace, said, “With ApertureDB, customers can quickly deploy, manage, and scale a data management platform on a reliable global infrastructure.”
5-2. Part of an ecosystem strategy
Google's strategy is not to control the AI market alone, but rather to build an AI ecosystem centered around Google Cloud. By incorporating highly specialized solutions such as ApertureDB as partners, we can provide more comprehensive AI solutions to client companies.
In fact, ApertureDB is designed to work closely with Google Cloud's generative AI tools and highly scalable cloud infrastructure. Client companies can launch ApertureDB in minutes from the Google Cloud console, and a 14-day risk-free trial is also offered.
5-3. Digital transformation support for companies
Mr. Vishakha Gupta, CEO of ApertureData, says, “We are excited to be able to utilize Google Cloud's cutting-edge generative AI tools and highly scalable cloud infrastructure.”
With this collaboration, companies can drastically shorten the construction of multimodal AI systems, which previously took 6 to 9 months. For Google, it is a strategic investment leading to long-term expansion of cloud usage by supporting the digital transformation of client companies.
6. How ApertureDB will change the future of business
The advent of ApertureDB has the potential to go beyond simple technological innovation and change the way business itself is.
6-1. Break down data silos
It fundamentally resolves the** “data silo” issue** that many companies have. Customer data from the sales department, campaign images from the marketing department, voice recordings of customer support, design drawings from the product development department, etc. will be integrated across departmental barriers.
As a result, cross-sectional analysis, which until now was difficult, can be performed on a daily basis, such as “utilizing customer voices in product development,” “formulating preventive measures from past trouble cases,” and “quantitatively measuring marketing effects.”
6-2. Creation of new business models
Data integration uncovers new business opportunities that were previously unseen. For example, retail stores may be able to develop a new business called an “optimal store layout design service” by integrating and analyzing in-store human flow data, product placement, and sales data.
In the manufacturing industry, new services such as “predictive maintenance services” and “customized products based on usage conditions” may be created by integrating product usage data, maintenance history, and customer feedback.
6-3. Democratizing AI for small and medium enterprises
Until now, the use of AI has been a monopoly patent for large companies. However, cloud-based solutions like ApertureDB make it easy for small to medium businesses to deploy advanced AI systems.
Utilization where the town's photo studio analyzes customer photo data and preferences and makes personalized album proposals, and local farmers integrate crop growth images with weather data to predict optimal harvesting times will become realistic.
An era where data is “connected” may have arrived
ApertureDB is a technology** that “connects” data that existed separately. Information in different formats, such as photos, videos, audio, documents, and numerical data, is managed in a single system, providing an environment where relationships with each other can be understood and searched instantaneously.
This isn't just a technological advance. It means the advent of an era where business decisions are made based on comprehensive data analysis rather than intuition or experience.
The technology “AI understands images,” which began with smartphone photo searches, has now evolved into a strategic tool that affects the competitiveness of companies. As a technology at the forefront of that evolution, ApertureDB will fundamentally change the way we work and the way we do business.
In an age where data is “connected,” ApertureDB is not just an option for companies, but may become an essential infrastructure for survival. That wave of change has already begun.
