ExaWizards' Engine for Solving Social Problems - Technology Management Department

Hello, my name is Tomohiko Kimura. I am the lead of the Technology Management Department. I want to introduce you to the Engineering Department, the engineering organization of ExaWizards.

What is our mission “solving social issues”?

First of all, if you are wondering what kind of company ExaWizards is, Mr. Oue's previous article tells you about the company as a whole. In a word, ExaWizards solves social issues through artificial intelligence for future generations.

What is exactly a social issue?

Specifically, we are tackling a wide variety of issues, such as DX challenges many Japanese companies face, nursing care issues in Japan's hyper-aged society, the impending workforce decline, and working Japanese people's mental health.

As you can see, ExaWizards is a large-ranged company that tackles a wide variety of areas and issues. We are issue-driven and business-driven, not technically a technology company, so we conduct various businesses ranging from DX, robotics, mental health, pharmaceuticals, and nursing care.

Technology Management Department is an enabler in terms of technology

The Technology Management Department copes with technical difficulties for advances in these diverse businesses.

For example, we provide a forum for companies to exchange information on DX, create content for training DX personnel and product managers, and support companies in their DX efforts. The Technology Management Department develops assessment and e-learning systems. The department is also developing SaaS products that allow users to create machine learning models with no code, without programming or specialized data science knowledge.

On the other hand, we are building an online platform for parallel and asynchronous processing of 1-on-1 online videos in the cloud, and developing various machine learning models that work in Docker containers on that pipeline.

We have also developed a product called Humanitude??, a training program for French caregiving techniques. The product enables Humanitude instructors to comment directly on videos of trainees providing care for making the instruction more intuitive. This product is now being developed as "TORTO," a gait analysis AI service, in collaboration with welfare equipment providers.

The term SDGs has become common. Energy challenges are especially a major social problem both in Japan and abroad.

Our machine learning engineers are tackling cutting-edge data science challenges such as optimizing power generation and forecasting electricity demand. We also build MLOps environments to realize integrated maintenance, operation, and continuous development of machine learning models in actual corporate production environments.

I briefly introduce the technical side of this company.

The Technology Management Department consists of five divisions: AI Frontier, AI Product Development, Edge AI, Platform Engineering, and Infrastructure & Security.

Social Issues and the Frontiers of AI - AI Frontier

The AI Frontier Department addresses verification testing and production operation of real problems that companies and organizations are facing, using AI and data science.

Almost all of the members are machine learning engineers. They are assigned to four technical groups: structured data, images, natural language processing, and mathematical optimization.

The department of AI frontier has many unique engineers from various backgrounds and countries.

Let me introduce my members. Mr.Satoma, Machine Learning Engineer in ExaWizards, leads the guild addressing structured data. He has an academic background in mathematics and physics, which allowed him to work on numerous R&D projects in Canon.?

Mr. Onishi leads the guild of natural language processing. He has been interested in language which is the main part of human beings since his childhood.

Mr. Imanaka is a tech lead in the guild of mathematical optimization.

AI Product Development - DX Products

Our businesses must be scaled to have a greater impact on solving social issues. The AI Product Development Department is responsible for creating products to achieve scale.

The problems of AI or data science are different from raw real problems. To solve real social problems is necessary to deal with the whole ecosystem around the problem. ExaWizards is developing a platform called "exaBase '' for solving DX issues for Japanese companies to scale their businesses.

The challenge of DX begins with people.

The exaBase platform starts with the "exaBase Community". The platform enables companies to collect information about DX. "exaBase DX Assessment & Learning" supports the discovery and training of internal human resources who can take on DX. "exaBase DX Learning" trains employees for their learning DX. exaBase Prediction and Analysis" and "exaBase FAQ" build AI systems, using in-house data. There are a series of products such as "exaBase 1on1" to help companies introduce and implement 1on1 more effectively as a tool for developing and motivating subordinates, and "exaBase Wellbeing" to help support the mental health of employees.

Ritsu Kimura, a winner of the Best Engineering Manager Award in the company's internal awards program, leads the Product Development Department. This department is responsible for the development of exaWizards' core DX products, such as exaBase Community and exaBase DX Assessment & Learning.

AI Product Development - Care Products

One of ExaWizards' product development goals is to create smarter human-to-human and human-to-system interfaces through the intervention of AI.

"Hanast" helps caregivers keep records and focus on more essential tasks in the area of nursing care.

"TORTO" makes video-based communication smarter, promoting collaboration among nursing care service providers, medical institutions, and welfare equipment suppliers that offer independent support services to the elderly.

Many of the team members who work in the area of nursing care, including the engineers, have a special passion for this field. According to Mr. Yuki, the product owner of "Hanast", everyone on the Hanast team is kind! (I don't think it means that other engineers are not kind, lol.) Mr. Kameyama, a machine learning engineer, has been the tech lead on this team since its inception, has now moved on to become a product manager, and is doing a spectacular job.

At the same time, a product should be designed to allow more users to accept the product without being aware of the technology or AI. The Design Group of the Product Development Department, led by Akasaka-san, the most passionate human designer at ExaWizards, is a group of designers who collaborate closely with the product development team.

AI x Hardware - Edge AI

The Edge AI Department aims to solve social issues through AI x hardware, including hardware as the interface between people and AI products and as the boundary of the ecosystem for social problems.

The Edge AI Department promotes the most technology seed-driven development in the Technology Management Department. The Robot AI Group, led by Mr. Asatani, has been developing multi-modal AI technologies that ExaWizards has created. Based on this, we are developing general-purpose gripping systems, powder weighing systems, and skilled skill visualization technology.

The team is also made up of many talented engineers, including team member Ceron-Lopez, who won the company's Best Wizard Award in October.

The AI Camera Group is also developing human tracking and object visualization technology using AI cameras designed and developed by ExaWizards, which can perform AI processing on the edge. The visualization technology detects foreign objects in vials containing vaccines, etc. Tracking technology is being tested in large warehouses and farms. We aim to commercialize the product.

Platform, Infrastructure, Security

As we have introduced so far, ExaWizards is simultaneously engaged in AI product development, demonstration testing, and production operation in various areas. In addition, the number of client companies continues to increase, while each of them has different security requirements and requires a higher level of security than before.

A large number of projects and clients with various systems and security requirements require us to have the platforms in a unified manner such as authentication platforms to develop products and services more efficiently and more securely.

The Platform Engineering Department, led by Kuti, has two groups: the Platform Development Group, which develops common platforms, and the Platform Engineering Group, which develops products and services. The Platform Operations group is responsible for the maintenance and operation of various machine learning models and services. The common platform is the foundation and infrastructure for the operation of products and machine learning models.

Mr. Scott, who leads the Platform Development Group, is developing the product infrastructure and society with a desire to update the "structure" of the product. On the other hand, the Platform Development Group, which is responsible for the operation of the product and machine learning models, is the lifeline of ExaWizards. The Platform Engineering Department is a gathering place for geeky engineers with a deep passion for technology. In his note, he will introduce more about the Platform Engineering Department.

Otherwise, the Infrastructure & Security Department, CISO of ExaWizards, is responsible for providing secure IT infrastructure and tools to all ExaWizards' employees. The Infrastructure group is responsible for managing internal IT systems such as Microsoft Teams and Microsoft 365 and various cloud service accounts, as well as conducting security reviews of the systems we develop.

AI Engineering Fellow and AI Technology Promotion Office

I have introduced the structure that supports the engineers who are creating AI at ExaWizards. ExaWizards has also AI Engineering Fellows (AIEFs for short) whose mission is to create the future of AI.

AIEF creates a vision for future AI technology and leads the way in spearheading R&D for advanced AI technology. In the team, Mr. Asatani started the Robot AI group in the Technology Management Department from scratch and established the foundation for ExaWizards' multimodal AI technology. Mr. Sudo is always full of unique ideas. Mr. Satoaki has steadily promoted the creation of good systems and developed the massively asynchronous parallel execution platform for machine learning tasks that supports "exaBase 1on1". Mr. Endo has used his coaching skills to create a flexible and autonomous organization of machine learning engineers.

They belong to the AI Technology Promotion Office, an independent and cross-functional organization of the Technology Management Department, and are working to create new seeds of technology for ExaWizards.

Our philosophy of software design and what we request in our members

Here is a brief overview of how the machine learning engineers and product development team develop their work.

The Technology Management Department is a flat and autonomous decentralized organization. For example, the lead of a product development team is called a "tech lead", but the role is like a mini-CTO of a startup. Tech Leads are expected to make plans, manage the team, and make proposals to the business side, while also anticipating the needs of the actual users, the market, and the business side.

We expect engineers other than tech leads to always think about what is best from the perspective of the business and the issue, as if each of them were the owner of the product or the founder of the company, and to step in and make proposals on the business side themselves.

Weekly management and business meetings are openly attended within the company. Minutes and recordings are always shared, so we always have access to what the management team is thinking right now. If an engineer feels that the company has something to improve organizationally, he or she can request a 1-on-1 with a director or the president to discuss it.

Styles of Development vary from department to department. For example, many teams operate according to general agile development practices. The business manager plays the role of the product owner and communicates the views of actual users and the business unit. The product designers participate in the support of the product owner. The team will consist of front-end engineers, mobile engineers, and designers.

On the other hand, machine learning engineers in the AI Frontier Department belong to each of four groups by technical area: structured data, image, natural language processing, and mathematical optimization. Each guild has several engineers called "leads," and the guilds function autonomously and flatly as a place for sharing and accumulating knowledge, collaborating with engineers, and building technical assets.

Conclusion

This article tells you about what the Technical Management Department is an organization that does, what kind of engineers it has, and what it does. ExaWizards is a company providing the business of solving social issues. In other words, it is a very issue-driven and business-driven company.

I believe that solving issues by utilizing AI means how to make people's interactions with the world, with systems, and with each other smarter. Therefore, the problem to solve comes in all directions. If we tackle them individually, it will be a tremendous war of attrition, and there is no way to win.

We need to technically select and focus on challenges coming from all directions. We must strategically conduct R&D, hire competent engineers, and partner where necessary. Ultimately, we believe that technology issues are also people’s issues.

Do ExaWizards' engineers find the issues, technologies, and products interesting, meaningful, and challenging? The objective of ExaWizards' Technology Management Department is to provide an environment for their passion.

With the company's listing, the Technology Management Department will be scaling even more aggressively. Whether you are a machine learning engineer, software/hardware engineer, or designer, we are looking for someone who strongly shares our mission to solve social issues by addressing AI and other software technologies.

We want to work with those who want to go beyond the framework of academia and demonstration projects, tackle real challenges, apply their technical and design skills to create an impact on the SDGs, or like developing and designing systems that are easy to use for people from many different backgrounds in many different areas.