Intelligence by Wizards from around the world. The social issues of a super-aging society, which Japan is experiencing first hand, will become issues for the world in the near future. The mission of ExaWizards includes the global deployment of the results of R&D conducted in Japan, a country with advanced issues. We have a very diverse group of people, such as AI researchers and engineers who are also university professors, unexplored creators, a former CTO of a startup that was bought out by Google, and a former owner of a professional baseball team, we have people with various backgrounds working together happily. From wizard-grade engineers to wizard-grade care gurus. Let’s gather together and create a better future.
roles exist in creating our own product - Definition of each role in ExaWizards Software Engineers Responsible for the development of our own services. Depending on the phase of the product, handle case definition and algorithm optimization, develop the front- end or the back-end. Machine Learning Engineers Responsible for implementation and improvement of learning models and algorithms. Also responsible for data analysis and product development depending on skillset. Platform Engineers Responsible for building and automating the development platform for our own products that are launching continuously. Also, in charge of cloud security, GPU environment, and other areas. Designers Responsible for product experience, interface, and visual design. Depending on the situation of the team, also responsible for front-end development. Product Managers Responsible for various tasks in launching and growing the product. Mainly responsible for drawing up the future of the product, prioritizing its execution, and assigning the appropriate members.
each role is involved in the two businesses • AI platform • AI product business Due to the characteristics of our products, at ExaWizards we add Machine Learning Engineer and Platform Engineer to ‘Product Manager + Software Engineer + Designer’ that are generally required for product development Software engineer Product manager UX Designer Machine Learning engineer Platform engineer Communicate directly with the client depending on the engineer's own orientation (If not oriented, AI consultant will be in charge). Project managers and UX designers who drive the progress also participate in the team. Each member can also communicate with the client depending on their orientation (omitted in the diagram). Client AI Consultant Machine Learning Engineers Project Team Large scale project teams Client AI Consultant UX Designer Project managers Machine Learning Engineers
have developed are implemented in society, you can feel firsthand that you are "solving social issues" and "contributing to the society.” There are many projects in different phases in a wide range of domains such as Care / HR / Fin / Med / Robot / VisionTech, etc., so you can be involved in project policy development and technology selection. Contributing to the resolution of social issues Discretionary involvement in a wide range of domains What attracts engineers to ExaWizards? About Engineering team at ExaWizards Able to engage in the development of services and products that can be used in the field, not only by using APIs for learning models and optimization algorithms, but also by incorporating applied technologies. Create services from zero/ Product development
benefit system: Doctoral course support program This system supports employees who wish to study the doctoral course of the Graduate School. Eligible fields of study: Research fields related to our mission Support contents: Financial support until graduation (entrance fee, tuition, etc., up to a total of 2,000,000yen) Work style: In addition to research as a working doctor, you can apply one day a week on weekdays to research at the university Who is eligible? 1.Full-time, contract employees, and executive officers of the company with a master's degree 2. Prospective employees who are currently enrolled in a doctoral course or planning to enter a doctoral course
benefit system: Self-learning support Study sessions: A 60-minute study session is held once a week. Presenters include machine learning engineers, software engineers, and members from the legal and business department, as well as university professors who are under contract for advisory. Book purchase and conference support: We support purchasing of books and papers related to your work, participation in academic conferences and events, and travel and accommodation expenses for these.
benefit system: Invention promotion system Incentives are provided at each stage of patent notification, application, registration and internal awarding. Both engineers and sales staff are encouraged to obtain patents, and dedicated officers in charge are in our legal team. In FY2019, we obtained 50 patents and won the Grand Prix in the start-up category of the IP BASE AWARD sponsored by the Japan Patent Office. The Japan Patent Office (JPO) is holding the first "IP BASE AWARD" to recognize the best startups, IP support experts, and ecosystem players who excel in IP strategy. ( https://ipbase.go.jp/news/2020/03/news-0319.php)
benefit system: Discretionary work system We have introduced a discretionary work system for engineering and design positions, and for some business positions. There is no fixed start or end time, and daily work hours are very flexible.
variety of club activities, including the Math Club and the Quantum Computer Club. Math Club Go /Shogi Club 3D Printing Club Quantum Computer Club Competitive Programming Club Football Club BabyTech Club Traditional Arts Club Animation/Manga Club For more information, please ask us in the interviews!
growth, and resolution of social issues. "I, You, We": An "autonomous team" that aims for the triple win Icons made by Pixel perfect from www.flaticon.com Divide into Teams according to the expertise Action Guideline Initiatives for career development Image Processing Structured Data Processing Natural Language Processing Mathematical Optimization Project assignments to other guilds as sub-members Study sessions such as lectures by prominent professors Regular 1-on- 1s with your supervisor Accumulating, organizing, and deploying technology stacks / Collaborate with other guilds and teams /Determine OKRs, such as creating projects that lead to career growth
Processing Guild ▪Guild overview • We build machine learning models for image processing such as image, video, and point cloud data in projects. ▪What kind of things does the guild do? • It handles all data related to images and point clouds. As long as it is an image, we handle both input and output. Semantic domain separation Object detection Defect product detection Project Examples Modeling the Distribution of Normal Data in Pre-Trained Deep Features for Anomaly Detection (arxiv.org) Segment road surface markings and buildings from aerial views, roads, and point cloud data. It can be used for road surface extraction. Detect people, animals, vehicles, etc. from retail stores, fixed cameras, construction sites, etc. It can be used to measure the number of objects, detect accidents, etc. GitHub - Nathan-UW/OpenGF: OpenGF: An Ultra-Large-Scale Ground Filtering Dataset Built Upon Open ALS Point Clouds Around the World Learn image features of good products by using image generation techniques such as GAN to classify defective industrial products and detect defective parts.
Processing Guild ▪Guild point - We have members who have been accepted to SIGGRAPH, CVPR, etc., and members who are involved in the management of international conferences and domestic research groups, so we have particularly active exchanges with academia. - You'll be exposed to images from many different fields! For example, you can build models of highly specific images such as medical images, industrial product images, and fixed camera images while discussing with professionals who have domain knowledge. - You can be actively involved in not only model building but also project design. - You can work on projects while discussing with high quality consultants - Our advisor is Dr. Hideki Nakayama, Associate Professor, Graduate School of Information Science and Technology, The University of Tokyo. ▪Image of future growth - We would like to build models using our own data and create new projects and products using Milcube (our AI camera). - In order to increase collaboration with academia, we would like to write more papers and make more external presentations. If you have the following expertise/background, let’s have a casual conversation with us! - Experience in building machine learning models using images, point clouds, etc. as input/output. - Those who like to study papers about CV, CG. - Experience in presenting at conferences and writing papers. - Experience with product integration, etc.
from the School of Advanced Science and Engineering at Waseda University, he entered the Department of Physics and Applied Physics at the university and completed the doctoral course. His major is 3DCG animation generation and image processing. During his studies, he was selected as a visiting researcher at UC Berkeley Artificial Intelligence Laboratory. He is also a part-time lecturer at Komazawa University's GMS Department and a visiting researcher at the Morishima Laboratory, Graduate School of Advanced Science and Engineering, Waseda University. He joined ExaWizards in April 2019. Takuya KATO Completed Master's program in Physics, Graduate School of Science, Osaka University, March 2014. Specializes in first-principles calculations. In April of the same year he joined Murata Manufacturing snd engaged in developing package materials for communication modules. In November 2017, he joined ExaWizards as a machine learning engineer, where he is involved in image/video related tasks. Akihiro FUJII AbdulRahman ABDULGHANI After graduating from the University of Fukui with a Ph.D. in engineering in 2014, he joined a research project at the Petroleum Institute in the United Arab Emirates in 2015, where he studied image classification using complex neural network deep learning of reflection seismic waves. In 2016, he joined the Tokyo branch of Universal Robots, Inc. as an R&D engineer, working on human palm spoof detection, and then joined Matsuo Laboratory (WEBLAB), University of Tokyo as a project researcher. He joined ExaWizards in 2018 as a Machine Learning Engineer.
Data Processing Guild ▪Guild Overview - Perform machine learning and data analysis on data in table format ▪What kind of things does the guild do？ - Mainly develop learning models that deal with data in table form. - They also work on a wide range of other projects that do not clearly fit into any of the natural language, optimization, or image guilds. Human Resources: Job offer acceptance forecasting Retail: Demand and sales forecasting CS: Complaint Forecasting Project Examples
Data Processing Guild ▪Guild Overview - We aim to strike a balance between business value and scientific/theoretical justification. - ExaWizards itself works with a variety of industries and companies, and we receive projects that are cross- disciplinary or not applicable to other guilds. This is a unique opportunity to be exposed to a wider range of technologies. - We actively attend external conferences and study groups as much as possible. - The members are diverse and have good relationships. - Our advisor is Professor Hisashi Kashima, Graduate School of Informatics, Kyoto University. If you have the following expertise/background, let’s have a casual conversation with us! - From a machine learning/statistics lab and have studied it extensively. Or you have PoC experience with other companies. - Able to look at your work critically (and proceed carefully). - Know or are trying to understand the working principles of the algorithms and libraries you use. - Willingness to learn domain knowledge. ▪Image of future growth - While implementing the latest research, we would like to cover the methodologies of machine learning and statistics that have been established to some extent. Beyond that, ideally, we would like to increase the number of people with knowledge in both theory and practice. - we would like to create small subgroups with different specialties and create a system that allows us to pursue both the breadth and depth of our techniques at the same time.
Engineering, Kyoto University. In his previous position at Canon, he was engaged in research and development of physical simulators for multi-component fluids and other phenomena analysis, various numerical calculations, image processing acceleration, and mixed reality system development. After joining ExaWizards in July 2017, he has promoted various prediction model building projects mainly using structured data, such as building a prediction model for izakaya sales and visualizing a prediction model for compound activity. Guild Members _ Structured Data Processing Guild Masanori SATO Graduated from the Graduate School of Engineering Science, Osaka University. Belongs to the Laboratory of Mathematical Statistics. Member of Japan Statistical Society. In his previous job, he was involved in the development of IoT devices, and after joining ExaWizards, he has been working on the following projects. - Shift adjustment in a restaurant chain, especially predicting the probability of employees' acceptance - Abnormal transaction detection, change detection, robust estimation in bank accounts He is particularly strong in the analysis of structured data. Hiroaki KOBAYASHI Keishiro KAWABATA Graduated from Kyoto University Graduate School of Engineering, Department of Materials Science and Engineering, and joined Exawizards in 2019. He was in charge of the development of an AI assessment model for used condominiums developed with Mitsui Fudosan Realty, as well as projects in a wide range of industries, including a medical data project with a chemical manufacturer and retail projects with drugstores.
Language Processing Guild ▪Guild Overview This is a guild that deals with text data. The main language we handle is Japanese. (There are a few cases where we handle English and Chinese). ▪What kind of things does the guild do? We develop learning models for projects and products that combine natural language processing and other technical domains. Relation extraction task Tasks for extracting relations from natural language (documents) and applying them to automatic construction of knowledge bases, etc. Aspect-based polarity analysis task Tasks to determine positive/negative/neutral from documents with indefinite perspectives and multiple polarities even in one sentence Icons made by Pixel perfect from www.flaticon.com Project Example
Language Processing Guild ▪Guild characteristics - We have elements of both implementation and research - We have members with experience in not only learning models but also social implementation (product development/start- up experience) - ACL, COLING 2020 accept - Our advisor is Professor Jun Suzuki, Graduate School of Information Sciences, Tohoku University If you have the following expertise/background, let’s have a casual conversation with us! - Experience as a researcher(project) - Product experience ▪Image of future growth - We would like to further strengthen the skill set of product development within the guild, and make it a guild that can accelerate the flow from project to product. - At the same time, we are also looking at building systems, infrastructure, and internal resources to automate and improve the efficiency of our projects. - Currently, we are in charge of some voice-related work because it is strongly related to natural language processing, but in the future, we would like to form a separate guild of voice experts.
Science at Nara Institute of Science and Technology. Later at Just System, he was involved in the launch of an AI-based listing ad management support service and the maintenance of morphological analyzers. He joined ExaWizards in 2019. Have engaged in the development of viewpoint-based polarity analysis from HR-related documents, skill extraction from work history, and structured data creation from speech information. Guild Members _ Natural Language Processing Guild Masaki ONISHI Graduated from the Graduate School of Informatics, Kyoto University. He was in charge of planning, development, and operation of data-driven policies in the area of daily consumption at Recruit. Later he joined ExaWizards, where he’s in charge of a wide range of projects related to natural language processing for independent administrative agencies and major telecommunications companies. Extensive experience in model building and operational design from the perspective of the field. Yuichiro MACHIDA Shun HASEGAWA Master's degree in Information and Communication Systems, Tokyo Institute of Technology. Currently enrolled in a doctoral course. After five and a half years of research on NLP in Okumura Laboratory, he joined ExaWizards in 2021 as a new graduate. His research interests include automatic construction of Japanese sentence summarization datasets using syntactic information from text, and automatic dialect translation techniques that are robust even with small datasets using deep learning. His papers have been accepted to ACL, the most difficult conference on NLP, and he received the Young Scientist Encouragement Award from the Association for Natural Language Processing.
Optimization Guild ▪ Guild Overview - Unlike other guilds that mainly deal with learning models, this guild deals with mathematical optimization. - Mathematical optimization is widely used in decision support, where it is expected to increase profits, reduce costs, and create efficient measures that humans cannot come up with. It is sometimes combined with the results predicted by learning models. ▪What kind of things does the guild do? Mathematical optimization is a problem-solving technique to find the optimal solution to a real-world problem where the data, causal relationships and rules are known, but no solution can be found. The guild develops algorithms for projects using continuous optimization and combinatorial optimization methods. Wood cutting proposal and actual operation Development of an app for crew and managers to efficiently submit and consolidate shifts and a model for efficient shift creation. Automatic creation of coal operation plans Project Example
Optimization Guild ▪Guild characteristics - We have members with competitive programming backgrounds who are well versed in computational quantities. - This is an environment where you can specialize in mathematical optimization. - The know-how of experts is often not clearly defined, and in many cases it is necessary to convert the know-how into mathematical formulas. However, it is rare that the specifications are fixed from consulting to delivery, so it is unique and important to be able to do the whole process from consulting to delivery. - Our advisor is Prof. Shunji Umetani, Department of Mathematical Optimization, Graduate School of Information Science and Technology, Osaka University. If you have the following expertise/background, let’s have a casual conversation with us! - Those who like algorithms and puzzles - Those who want to intervene and solve customer problems directly ▪Image of future growth The number of members is still small, and we are actively hiring talents. We do not yet have a system for sharing and accumulating know-how, so we would like to build it in the future.
Graduated from Kyoto University, Graduate School of Informatics. After working on system development at a SIer, he joined ExaIntelligence (now ExaWizards) in 2016. At ExaWizards, he is the PM for Yoshinoya x SECOM's AI for creating shift schedules for part-time, and has engaged in many other AI and optimization projects. He is in charge of the algorithm of HR haichi-kun. Takeshi IMANAKA He specializes in discrete optimization. After graduating from the Faculty of Science, Kyoto University, he received his Master's degree from the Department of Mathematics, Kiel University. At ExaWizards, he is a member of the development team of HR haichi, and has been involved in several optimization projects such as power generation plan optimization for electric power companies and shift optimization for logistics and restaurant industries. Yugo ISHIMARU Yuta KOHNO Graduated from the University of Tokyo, majoring in electrical engineering. After working on the development of automated parking systems at a Japanese manufacturer, he joined ExaWizards in September 2019. After working on product development as a software engineer for about a year, he is now mainly in charge of mathematical optimization.
solutions. exaBase One of Japan's largest platforms for leads to success with AI and DX ※exaBase, CareWiz, CareCoach, and Torinku are registered trademarks of ExaWizards. AI SaaS products developed in-house using exaBase's technology components to solve social issues such as elderly care and childcare. An auto photo shooting and sorting application for childcare sites. Supports productivity improvement in childcare settings. Realize Science-Based Care with AI x Care Visualizing gait using AI and skeletal posture estimation to help improve the quality of care recommendations. Video guidance on how to provide care in specific situations, and AI supports search for local information. Support solving problems in family care giving. Simplifies care staff's indirect tasks, such as recording and sending off care contents, by using AI and voice input to support more efficient operations. … Prediction & Analysis Robotics Corporate Search Edge Camera Well-being Community DX Assessment & Learning Abnormal transaction detection Hanast Toruto Kazoku-shien
products that contribute to digital talent development and DX promotion 1. Product Overview 2. The role of the engineers Data analysis/prediction tool that anyone can use easily without programming. Can be used in many industries. ＜ How to use > AI learns, extracts features and analyses the data in your company. • Development of new functions • Maintenance (responding to improvement requests from customers, fixing problems) • Technology selection and feasibility study for new functions and algorithms • Development of new application products (SaaS) / design of system architecture for services Prediction & Analysis
People we want to work with 3. Meet the people you will work with • Those who can think from both technical and user perspectives and take proactive actions • Those who are passionate about maintaining code quality and establishing development processes • Those who want to write "I truly popularized AI" in their career • Those who have a strong engineering background. • Those who want to be part of an international environment career Born in Hong Kong. Graduated from the Faculty of Mathematics, The Chinese University of Hong Kong. After working as a web developer, he worked in the IT department of CPCE PolyU since 2013, mainly responsible for the implementation and maintenance of the HR attendance management system. He joined exaWizards in 2020 and is currently a software engineer for exaBase Prediction and Analysis. WONG Hang career Majored in Computer Science at Hong Kong University of Science and Technology and Built a workplace management system from scratch for a start-up company in Hong Kong. In 2016, he moved to BizReach in Japan, where he was involved in the development of a recruitment management system called HRMOS Recruitment. In October 2019, he joined ExaWizards, where he is currently the Tech Lead for exaBase Prediction and Analysis, and is also responsible for developing the exaBase Community. CHAN Wingkit
productivity improvement, care techniques, and mental health in clinical and nursing care environment. 1 .Product Overview 2. The role of the engineers This is an app that uses voice x AI to improve operational efficiency at care facilities and create more time for staff and patients to interact. ＜How to use＞ AI listens only words related to care records from speech and links the records to care report software. • System improvement and implementation of new functions through agile and scrum development • Consideration of improvements and new features with the business and engineering sides • Possibly be involved in development other than main responsibilities (front-end, back-end and mobile) • Depending on your preferences, do client's front-end work, product planning and operation, etc. Speak AI analysis Record CareWiz Hanast
to work with • Those who are Interested in mastering broad technologies, not just in one area. • Those who understand and can operate agile/scrum development processes. • Those who enjoy communicating with the clinical environment, as we often work with nursing care and medical institutions. • Those who are considerate to elderly and care staff’s feelings. • Those who can flexibly respond to changes in Japan’s nursing care system and national standards to develop products. career He joined ExaWizards in October 2020 after working as CTO of Chintai Joho Inc. After serving as Tech Lead for FinTech products, he joined CareWiz team in July 2021, mainly in charge of backend development. Kenji HARA career He has been working as an infrastructure SWE since the days of Digital Sensation, the predecessor of ExaWizards, and is in charge of the development of CareWiz, Toruto, and MedTech. He also works as an infrastructure and information security engineer. Yukimune KUROKI career After working in R&D and product development of image processing at a heavy electric manufacturer, he joined ExaWizards in November 2017. He has participated in several projects and products as MLE or SWE, and is currently the Tech Lead for CareWiz Hanast. Satoru KAMEYAMA 3. Meet the people you will work with
information to be extracted, reducing operational costs and security risks. Edge AI cameras can be widely used in various industries. exaBase Edge Camera Team Social implementation of a multifunctional portable AI edge camera that covers everything from human/goods flow detection to skeletal estimation -General service / system operations associated with service provision -System design for services -Development of hardware such as AI cameras and server-side systems -Adding functions and improving quality of services -and more. 1 .Product Overview 2. The role of the engineers Shooting video Extract necessary information Encrypt and send to server Museum Store Kids' spaces Hospitals and nursing care Workplace Sports
of change and do not hesitate to enjoy changing roles. • Have an interest in acquiring various technologies • Able to work with an ownership mindset • Those who pursues software quality • Knowledge of AWS and NVIDIA Jetson is preferred. exaBase Edge Camera Team Career Completed the doctoral program at the Graduate School of Systems Information Engineering, University of Tsukuba. He was involved in the renewal of the infrastructure system of a major blogging service, where he developed an image processing system. Later, he was involved in the launch of an IoT platform linked to a messaging service, and led the development of platform design, device design, and server-side systems. He is also in charge of developing new services linked to hardware at ExaWizards. Yohsuke FURUTA Career After graduating from Denver Metropolitan State University, he started working as a software engineer in Japan. Worked on various software projects from web, mobile, and backend systems to AI- related applications at the University of Tokyo. Currently, he is responsible for developing software services that run on a hybrid of edge AI devices and cloud infrastructure. Michael BAWIEC 3. Meet the people you will work with 4. People we want to work with
that could not be implemented by programming Weigh out powders/liquids Polish easily Damaged objects Assemble objects that require fine force control exaBase Robotics Team Social implementation of multimodal AI robots to support operational efficiency, mainly in the manufacturing and logistics industries Grasp food, boxes, and other irregularly shaped objects Spread and applying liquids • Robot system integration • New AI technology development • System maintenance and operation • Overall system design to connect robots and AI 1 .Product Overview 2. The role of the engineers Robotics
enjoy new challenges, including unprecedented ones. • Those who are interested in developing new AI technologies • Those who can handle system stabilization and infrastructure development • Those who want to be in an international environment Arturo CERON-LOPEZ Yuta MIYAZAKI Career Secretary of the Artificial Intelligence Research Group and head of the ExaWizards Engineering Promotion Office. He launched the AI Research Group in December 2015 while he was still a student at Osaka University. In June 2016, he joined ExaIntelligence Corporation (now ExaWizards) as an AI engineer. He is currently working on technology development focusing on system development that integrates robotics and AI technology, and was appointed as AI Engineering Fellow in April 2021. Satoshi ASATANI Career B.Sc. in Mechatronics from Universidad de Monterrey, M.Sc. and Ph.D. and M.S. degrees in Mechanical and Space Systems from Tokyo Institute of Technology and a research assistant in Shape Memory Alloys at HTWG Konstanz. In addition to research and development of mobility control and image processing algorithms for infrastructure inspection robots at HiBot, he developed middleware and user interfaces. Currently working as a TechLead for a robot vision product based on synthetic data. His hobbies include computer graphics and international exchange. 経歴 He received his M.S. degree from the Graduate School of Information Science and Technology, Osaka University in 2018. His major is computer science with a focus on image processing. Since 2020, he has been working at ExaWizards, where he is involved in a wide range of tasks, including API development for peripheral devices and robot motion generation using deep learning. Currently, he is working hard on a powder weighing system. 3. Meet the people you will work with 4. People we want to work with
of cloud infrastructure and common infrastructure for continuous product development 1. Department Overview Platform Operations Manage cloud infrastructure development and cloud security for the overall product Platform Development Develop a common infrastructure for APIs and software to facilitate agile product development • Common development environment / Cloud infrastructure development and maintenance • Cloud Security • Deployment of development tools • Development of common APIs for in-house products • Development of APIs for clients • Development of common infrastructure platforms and microservices common API Video Analysis Text Analysis Optimization Natural Language Processing (Example) (Example) Company A Product A Company B Product B Feature analysis and prediction Prediction & Analysis Well- being Hanast
• Those who can proactively challenge new things • Those who can involve others through polite communication • Those who are interested in utilizing a wide range of technologies Platform Engineering Department Career He has 15 years of experience in Systems Engineering, Web Design, DevOps, Infrastructure and Security. He joined ExaWizards in 2018 after managing as Global Head of Development Operations. Currently, he’s the Head of Platform Engineering. Patrick O. S. Kuti Career Graduated in Computer Science from the University of Adelaide. Worked in corporate and government security for about 8 years and joined ExaWizards at the end of 2019. Currently, he is a group leader of Platform Operations Group. Career After working for a telecommunication company where he was responsible for network operation design and operation tooling and development, he joined ExaWizards in 2020. He is currently in charge of architecting and developing common infrastructure in the Platform Development Group. Yusuke AOKI Lachlan MITCHELL 3. Meet the people you will work with 4. People we want to work with