Supporting data use in the AI era through synthetic data, annotation, and data analysis. Elith Inc. (head office: Bunkyo-ku, Tokyo; Representative Director, CEO & CTO: Koki Inoue; hereinafter "Elith") announces the establishment of a subsidiary, Elith Data Intelligence Inc. (hereinafter "EDI"), specializing in the design, creation, and evaluation of data needed for AI and robot development, as well as in the analysis and use of corporate data. EDI will operate its business around two pillars: "creating" the data needed for AI and robots, and "making use of" the data that companies hold. In addition to conventional data annotation, this includes generating synthetic data using AI, designing evaluation datasets, developing data for robots, and analy
News & other.
Official announcements and notes from the field and research, gathered across themes.
Related content
Elith Inc. is pleased to announce that the team including our Principal AI Engineer, Isaka, placed 4th and won a gold medal in the Kaggle competition "AI Agent Security - Multi-Step Tool Attacks," hosted by OpenAI, Google, and IEEE.
Supporting the development of talent who can not only "use" AI but also understand how it works and evaluate and implement it appropriately. Elith Inc. (head office: Bunkyo-ku, Tokyo; Representative Director & CEO: Koki Inoue; hereinafter "Elith") announces that, as an accredited program for the "E Certification" administered by the Japan Deep Learning Association (JDLA), it will launch the corporate AI talent development course "E Certification Prep Course Designed Backward from Real AI Development Work" from September 2026. This is the first course to be accredited by Elith as a JDLA-accredited program. While systematically learning the deep learning theory and PyTorch implementation needed to take the E Certification exam, participants will aim to acquire the "ability to appropriately evaluate models and outputs" and "the ability to understand how AI wor
Corporate AI use has already moved past the stage of debating whether to use it at all. AI features are built into more than just generative AI like ChatGPT and Claude — they're now embedded in the SaaS products and work tools people use every day, and the number of moments when employees use AI without even realizing it keeps growing. At the same time, not many companies have a real grasp of which AI is being used, by whom, and how, within their own organization. As convenience takes the lead, there are cases where awareness of risks such as information leaks, misuse, and accountability hasn't caught up. Amid this situation, C Certification is one entry point for building an AI governance system. So why has Elith positioned support for obtaining C Certification as a business, and what possibilities does it see the
Obtaining C Certification requires taking stock of AI usage, organizing risks, developing rules, and reviewing internal structures. Even after certification, reviewing AI usage and improving operations continues. That is exactly why this is an area where, if you try to go it alone, it's easy to get lost on questions like "where do we start," "how far do we need to go," and "how do we operate after certification." Elith can provide end-to-end support, from building AI governance before obtaining C Certification, to improving operations afterward, and even implementing AI security using GENFLUX Security. This time, we spoke with Shinomiya, who is involved in supporting C Certification, about how Elith positions "obtaining certification" and what kind of support it envisions beyond that. When it comes to supporting C Certification,
General-purpose AI like ChatGPT and Claude is one thing, but AI capabilities are now quietly built into the SaaS products and everyday work tools people already use — that's the reality of business in 2026. And it's no longer limited to specialists: it's now common for non-engineers to build their own work tools and automation flows using GitHub Copilot or generative AI. As this convenience grows, new challenges are emerging for companies. It's becoming harder to see the full picture of AI use — not just the AI a company formally adopts, but also the AI employees use on their own judgment, AI features added later to existing tools, and automation flows that have naturally taken hold in day-to-day work. What companies are being asked today isn't whether they use A
Supporting development capability in the age of AI agents together with governance and compliance in one integrated approach. Elith Inc. (Headquarters: Bunkyo-ku, Tokyo; Representative Director: Koki Inoue) is pleased to announce the establishment of "Elith AI Governance Lab," a specialized organization that ensures the safety and reliability of corporate AI adoption. ■ Background of the establishment The corporate adoption of generative AI and AI agents is rapidly shifting from the proof-of-concept stage to production operation. At the same time, more and more companies are facing challenges in ensuring quality and managing risk due to the non-deterministic nature of AI, and how to balance development speed with safety is becoming a management issue. Elith has to date developed the AI safety platform "GENFLUX
Elith Inc. (Headquarters: Bunkyo-ku, Tokyo; Representative Director & CEO: Koki Inoue; hereinafter "Elith") has launched "GENFLUX Security," a shadow AI countermeasure SaaS that helps enterprises visualize, control, and record their generative AI usage. GENFLUX Security is a security service that addresses the risk of "shadow AI" — employees using AI tools not approved by their company for work — and establishes an official route for enterprises to safely leverage generative AI. Service LP: https://lp.genflux.jp/ Background: The expansion of generative AI adoption and shadow AI risk While the business use of generative AI is rapidly expanding, on the enterprise side
Elith Inc. is pleased to announce the launch of "ELITH LOCK: FDE Challenge," a selective internship program for engineering students who work on product development close to client challenges. This program is not a mere development internship — it is a hands-on development program where participants can experience the full cycle from understanding client challenges to requirements organization, implementation, operation, and product improvement. In addition, for participants recognized for high practical contributions, we offer the "FDE 10M Offer," a starting salary of 10 million yen for new graduates. ■ What is ELITH LOCK: FDE Challenge ELITH LOCK: FDE Challenge is a selective program for students who, while understanding clients' business challenges, are involved in developing and improving Elith's own pro
Elith Inc. launches a series of practical generative AI courses for businesses and local governments
Elith Inc. is pleased to announce the release of three new practical courses to support businesses and local governments in embedding the use of generative AI into their operations: "Introduction to generative AI operating patterns for frontline work," "Introduction to AI governance and regulation for local governments," and "AI business creation and hypothesis validation course for new business staff." Rather than simply teaching how to use generative AI tools, this series is a practice-oriented program that focuses on how to safely incorporate AI into actual operations and keep it running continuously. In recent years, as the use of generative AI has rapidly advanced, businesses and local governments have faced challenges such as: not knowing how to apply it to frontline work; getting stuck at the PoC stage without adoption taking hold; concerns about information leaks and governance; and unclear rules and division of responsibility for AI use. AI
Elith Inc. announces that it has strengthened and systematized its corporate AI education program offerings. While the adoption of generative AI is advancing rapidly, many companies face challenges such as "having introduced AI tools but failing to establish their use in the field," "a shortage of AI talent," and "insufficient internal understanding of what AI can actually do." Building on insights gained through AI research and development, Elith offers hands-on AI education courses for executives, business staff, engineers, and researchers. Rather than stopping at mere knowledge acquisition, we aim to cultivate "AI skills usable in the field" and support companies in internalizing AI capabilities and transforming their operations. While AI adoption is advancing, “sustained adoption” is becoming a new challenge In recent years, an increasing number of companies have been advancing the adoption
Elith Inc. has launched "AI Safety REPORT by Elith," an owned media outlet that shares insights on the practical use and safety of generative AI. Under the concept of "The complete record of AI security that keeps Japanese business running," it shares information usable in the field, from implementation methods such as RAG (Retrieval-Augmented Generation) to quality evaluation and risk countermeasures. ▼AI Safety REPORT by Elith https://media.elith.ai/ This media outlet focuses on "implementation" and "safety" in generative AI adoption, providing not only explanations of how things work but also information directly relevant to practice, including deployment design, operational challenges, failure cases, and cost structures. In recent years, while the use of generative AI has been advancing, internal
Elith Inc. will run a mobile advertising campaign in the Shibuya area on Wednesday, April 1, 2026. On that day, for three hours from 18:00 to 21:00, nine advertising panels will move through the streets, directly approaching passersby. This is a unique campaign using a walking-type advertising method. Please keep an eye out when you're in the area. [Overview] Date and time: Wednesday, April 1, 2026, 18:00–21:00 Location: Shibuya area Ad format: Mobile advertising (9 units, walking-type) In addition, in conjunction with this advertising campaign, we will grant free credits for "GENFLUX," which lets anyone easily develop AI apps. [How to claim the coupon] ① Access GENFLUX (https://app.genflux.jp/home)
Elith Inc. is pleased to announce that "How Large Language Models Work: Learn by Building One," a book to be published by Nikkei BP on March 20, 2026, has achieved the following rankings on Amazon Best Sellers (as of March 17, 2026). "Introduction to Programming" category: No. 1 "Software Development & Languages" category: No. 1 "Development Techniques" category: No. 2 The book takes the large language model (LLM) "GPT," the foundational technology behind tools such as ChatGPT, as its subject, systematically explaining everything from the basics of the Transformer to the implementation of a GPT model. Through PyTorch code and diagrams, readers can understand how modern AI technology works by getting hands-on. It also covers alignment techniques (SFT, D
Elith Inc. will hold a seminar event at Shosen Book Tower (Akihabara) to commemorate the publication of the book 'How Large Language Models Work: Learn by Building One' (Nikkei BP), on sale March 20, 2026. ▼Details and registration here: https://connpass.com/event/387948/ This technical book lets you systematically learn everything from the basic mechanisms to the implementation of "GPT," the large language model (LLM) used by ChatGPT, by building it from scratch. It covers the fundamentals of the Transformer, building a GPT model, the mechanics of alignment (SFT and DPO), and even the latest reasoning-enhanced models. At the event, the authors will discuss behind-the-scenes stories from writing the book and technical details, as well as the latest developments in LLM
Elith Inc. will put up station advertising at Kudanshita Station (Tokyo Metro / Toei Subway) from Monday, March 23 to Sunday, March 29, 2026. This ad will feature posters carrying our company's message displayed inside the station. If you use Kudanshita Station, please take a look. *Please refrain from asking station staff about this. Display overview Location: Kudanshita Station (Tokyo Metro / Toei Subway) Display period: Monday, March 23 – Sunday, March 29, 2026 #ElithのAI伝言板 *The display period is subject to change without notice.
Elith Inc. will publish, through Nikkei BP, "How Large Language Models Work: Learn by Building One" (「作ってわかる大規模言語モデルの仕組み」), a book that lets readers learn the mechanics of large language models (LLMs) by implementing them, on March 20, 2026. Using GPT, the LLM that underlies technologies such as ChatGPT, as a case study, the book systematically explains everything from the fundamentals of the Transformer to the implementation of a GPT model. Using working PyTorch code, readers can understand the structure and training methods of LLMs. It also explains alignment techniques for generating responses aligned with human intent (SFT and DPO), as well as reasoning models such as Chain-of-Thought. It further introduces technologies involved in real-world LLM development, such as distributed training across multiple GPUs
Elith Inc. announces that it has entered into a business alliance agreement with AlpacaTech Co., Ltd. (head office: Chiyoda-ku, Tokyo; Representative Director & CEO: Morifumi Yotsumoto; hereinafter "AlpacaTech") to provide AI diagnostics for financial institutions and other organizations using AI guardrail and AI security technology. Through this alliance, the two companies will jointly study and develop services aimed at ensuring safety and reliability in the use of generative AI. Background and purpose of this business alliance As the business use of generative AI spreads, risks such as inappropriate output, information leaks, and governance/compliance response have become challenges. Financial institutions in particular are required to maintain a high level of safety and accountability when using AI. AlpacaTech is a financial
Elith Inc. recently welcomed a professor (Dr.) and about 10 students from Addis Ababa University in Ethiopia for a technical exchange. This visit was carried out to deepen understanding of trends in AI research, the potential for real-world deployment, and the future shape of international collaboration. On the day, the professor and students from Addis Ababa University showed great interest in Elith's research, development, and business. In particular, there was lively discussion centered on expert, practical questions spanning both research and implementation, including design trade-offs in multi-agent AI, scaling strategies for AI-driven businesses, and the potential for AI applications in agriculture and healthcare
Elith Inc. announces that its employees achieved top rankings in the data analysis competition "#23 Turing × atmaCup 2nd" (held January 16–25, 2026), organized by atma Inc. and co-hosted by Turing Inc. This competition was open to the public, with 282 teams participating and 2,216 prediction submissions made, making it a highly competitive event. Among them, our employees achieved the following results. ■ Results 5th place: Nagasawa 11th place: Fujihara Both are top-tier results achieved among a large field of participants, demonstrating the high level of technical skill Elith has in the fields of data analysis and machine learning. El
Lately, we often hear stories like "AI got it wrong" or "AI gave a strange answer." Did you know that even convenient AI carries risks of misinformation and information leaks? Attracting attention amid this is GENFLUX, developed by Elith. This service acts like a "health checkup tool" that checks the accuracy and safety of AI's answers. It detects errors and dangerous outputs from AI, helping businesses and local governments use AI with peace of mind. On the program, CEO/CTO Koki Inoue said, "If we just manage the risks, AI will move society steadily forward." GENFLUX is a technology born in Japan to create a world where AI and society coexist. See GENFLUX detai
Elith Inc. will hold an online webinar themed on AI adoption in construction and manufacturing, "Construction & Manufacturing × AI Frontline — The Next Competitive Advantage Seen from On-Site Implementation," on Wednesday, January 28, 2026. In construction and manufacturing, while the adoption of AI technology is progressing, many cases remain stuck at the PoC stage without reaching on-site implementation. In this webinar, through a dialogue-style panel discussion between Elith and Lightblue, we will explain the main tasks and AI technologies in construction and manufacturing, domestic and international adoption trends, and the factors behind PoCs stalling and their solutions, from both an implementation and a business perspective. With an eye toward on-site implementation, SaaS adoption, and joint research, the session presents points to consider going forward and practical appro
To strengthen its business strategy and real-world deployment structure in the AI and advanced technology domain, Elith Inc. is pleased to announce that it has welcomed Ryutaro Okada, who has served in roles including Executive Director of the Japan Deep Learning Association (JDLA) and Secretary-General of the AI Robot Association, a general incorporated association, as a management advisor. Background to the appointment as management advisor Elith has to date supported clients end-to-end from research and development through to real-world deployment, centered on generative AI and machine learning, across a wide range of fields including manufacturing, finance, healthcare, mobility, and disaster prevention. In recent years, AI technology has been shifting from the phase of "can it be developed" to the phase of "how can it be made to take root in society and continuously generate value." Okada, in addition to his experience in startup
Elith Inc. is pleased to announce that it will support, as a co-hosting company, the review session for the "21st atmaCup" data competition hosted by atma Inc. This review session will look back on the content of the recently held "21st atmaCup," providing a venue for participants to deepen their learning through solution walkthroughs by top-ranked winners and lightning talks (LTs). Anyone interested in the content is welcome to attend, even those who did not participate in the competition. Highlights of the review session: Gain in-depth insight into the competition's results Top-ranked winners' solutions and approaches will be introduced, with opportunities for discussion and questions among participants. Lightning talk (LT) presentation slot
Elith Inc. announces that it will support "atmaCup 21st Competition," a data competition hosted by atma Inc., as a co-hosting company. This competition will be held as a prompt battle-style competition, testing the prompt-design skills of both "attack" and "defense" — a new evaluation axis for the generative AI era. Since participation only requires submitting prompts, even first-timers can take part with confidence. ■ What Is atmaCup? A data competition operated by atma Inc., highly popular as an event where participants analyze problems in a short time and compete on accuracy. For this 21st edition, Elith is joining as a co-hosting company, offering new learning and competitive experience
Elith Inc. CEO & CTO Koki Inoue spoke at the webinar "AI Data Strategy for Balancing AI Performance and Safety," hosted by the AI portal media "AIsmiley" on Tuesday, November 18, 2025. The webinar introduced data strategies and risk management methods that companies can put into practice, focusing on the themes of "quality control" and "safety," which are becoming challenges as the use of generative AI advances. Elith used its own AI risk management tool, "GENFLUX," to explain, with concrete examples, how to check AI behavior and suppress bias and inappropriate outputs. Practical methods for using AI safely and effectively, and key poin
Koki Inoue, Representative Director, CEO & CTO of Elith Inc., gave a talk for the JFE Engineering Labor Union, on the theme “The Changing Trajectory of AI: Technological Innovation and Future Outlook for Generative AI, 2022-2026.” In this talk, he looked back on the trajectory of generative AI, which has developed rapidly since 2022, and introduced, from multiple angles, its expanding use in society and industry along with future possibilities. As the pace of technological innovation accelerates year by year, the talk offered an opportunity to consider how to understand and make use of the changes brought about by generative AI. We hope that this talk will help union members grasp the new future that AI will bring. [About the JFE Engineering Labo
Elith Inc. has published an interview article about an AI course delivered to RYODEN Corporation. RYODEN operates across a wide range of fields, including manufacturing, electronics, and smart agriculture, and has been actively working to raise AI literacy in-house and promote digital transformation...
Elith Inc. will officially launch "GENFLUX," a quality evaluation platform that checks the response quality and behavior of generative AI in real time and automatically proposes improvements, from August 18, 2025. Background and challenges...
Elith Inc. has published an "AI Deployment Case Studies" collection on its official website, systematically compiling practical use cases from across industries drawn from the AI deployment projects it has worked on. The collection introduces concrete examples of AI adoption across a wide range of fields, including...
Elith Inc. has published a note article focusing on the reality of remote work and internal communication. This article closely follows a day in the life of a fully remote Elith employee. Clocking in on Slack, team morning meetings, a culture of reactions, document sharing—it introduces, through employees’ own voices, how Elith puts "high-density communication" into practice. It’s also a piece that offers a glimpse into Elith’s distinctive culture, including the lives of members living outside major cities and the ways the team stays connected. Elith will continue to regularly deliver note content that shares the reality of our ways of working, organizational culture, and technology development. We hope you’ll check out the latest updates and get to know the people behind Elith
Elith Inc. has published a new special page, "Co-creation Business & IP Licensing," which draws on its AI research and implementation expertise to support new business creation and the utilization of intellectual property through co-creation with companies and research institutions. This page systematically introduces initiatives that go beyond PoCs and contracted development, such as business co-creation and IP licensing. Reflecting Elith's stance of "AI built together," it also features co-creation case studies such as startup support and industry-academia collaboration. ■ Introducing the main content Why co-create with Elith Introducing the technical capabilities and flexible structure that support you from research through implementation and business launch Introducing co-creation schemes Business co-creation: joint development of new products and services IP licensing: providing proprietary models and systems
Elith Inc. is pleased to announce that a paper by our researcher Shimomura, “OD-RASE: Ontology-Driven Risk Assessment and Safety Enhancement for Autonomous Driving,” has been accepted at “ICCV2025 (IEEE/CVF International Conference on Computer Vision 2025)*,” a top international conference in the field of computer vision. The accepted paper “OD-RASE: Ontology-Driven Risk Assessment and Safety Enhancement for Autonomous Dri
Elith has introduced state-of-the-art GPUs (Graphics Processing Units)※. This significantly reduces AI model training and inference time, accelerating both AI development and real-world deployment. ■ Background: Truly driving the two wheels of “research x implementation” Elith works across a wide range of AI domains, including healthcare, manufacturing, and operational optimization. In recent years, as the computational resources required for improving model accuracy have grown, we recognized the need for a more powerful development infrastructure. With this background in mind, introducing GPUs has significantly reduced the processing time required for model training, inference, and evaluation. This is expected to shorten the lead time from PoC to production implementation, making the real-world deployment of research outcomes...
Elith announces that it has renewed its corporate logo with the aim of strengthening its brand image and enhancing corporate value. ■ Background to the logo renewal Elith has adopted "Accelerate" as its corporate slogan, and as a tech company that creates value precisely amid change and chaos, has taken on the challenge of solving a wide range of problems using technology centered on AI and creativity as its weapons. This logo renewal aims to more intuitively express Elith's MVV (Mission, Vision, Values) and to be reborn as a symbol of Elith today and the Elith to come. ■ The thinking behind the new logo The new logo shines as it transforms, accelerates, and eventually becomes a star — that kind of energy
We would like to express our heartfelt thanks to everyone who regularly watches and supports our channel. Elith's YouTube channel (@elithofficial) publishes videos focused on the business use of generative AI tools such as ChatGPT....
Thank you very much for always visiting Elith Inc.'s website. We have now published a new page dedicated to AI development. This page clearly introduces support for leveraging cutting-edge technologies such as AI agents and image AI, which are among Elith's specialties, as well as our end-to-end development process from PoC through on-site adoption and operational embedding. We have also included case studies based on actual deployments. ■ Introducing the main content Why Elith is chosen Introducing the technical capabilities built through research and the consistent structure that supports you from planning through operation Key technology areas AI agents: analysis / optimization / RAG / voice and text generation Image AI: anomaly detection / object detection / image OCR
Elith Inc. is pleased to announce that it has formulated "Elith Way." This guideline is a code of conduct that helps all Elith members reflect the company's Mission, Vision, and Value in their daily decision-making while acting ethically and responsibly. Going forward, Elith will continue to drive sustainable and innovative corporate activities centered on technology and integrity. [Background] Our company has set forth the mission of "remaining a tech company that gathers diverse 'particles'—people, organizations, and technology—and accelerates step by step through dynamic interaction and scattering," pursuing both technological innovation and real-world deployment. Realizing this mission requires value creation that goes beyond
Elith Inc. has newly established "MU LABs," an in-house research and development organization aiming to achieve both the real-world deployment of AI technology and technological innovation. In addition, Elith has completed formal registration with the "Research Institution (Researcher Resource Assurance System)*" under the Ministry of Education, Culture, Sports, Science and Technology, the Ministry of Health, Labour and Welfare, and other relevant ministries. Under this publicly recognized, trusted research framework, we will pursue further technological development and social contribution. ■ About MU LABs MU LABs is a lab-style research and development organization established within Elith Inc. Starting from real, on-the-ground challenges, it aims to expand the possibilities of AI technology through development that emphasizes deep technical exploration and reusability. MU LABs currently centers its activities on the following three pillars: Products
We would like to express our heartfelt thanks to everyone who regularly watches and supports us. On Elith's TikTok account (@elith892), we share relatable content centered on practical ways to use generative AI tools such as ChatGPT, along with test cases and everyday tips! We've received feedback like "this looks applicable to my work" and "this is useful in practice," and since launching the account in April 2025, we've earned a total of over 27,600 "likes"! We will continue to bring you hints for using AI in more accessible and concrete ways. Please take this opportunity to check it out. ▶ https://www.tiktok.com/@elith892
Making AI adoption more accessible. We have renewed our content. Elith Inc. is pleased to announce that its official YouTube channel has surpassed 1,300 subscribers. We would like to sincerely thank everyone who regularly watches and supports us. Elith's YouTube channel has mainly shared information on AI technology and use cases, but we have now renewed part of our content. We will deliver practical, approachable content that helps more people "concretely understand how to use AI and want to try it." Going forward, we will continue to clearly deliver, through videos, hints for using AI at work and in daily life, so we hope you will continue to
Elith is pleased to announce the publication of an interview article featuring Sugigaki and Makita of the Research Center of Engineering Innovation, Technology Headquarters, and Kobayashi of the AI & Software Development Department, DX Promotion Center, DX Headquarters, as a case study of hands-on support provided to JFE Engineering Corporation. Elith Inc....
Elith announces the publication of an interview article with Representative Director and President Junya Shibata as a case study of hands-on support for improving the efficiency of home-visit nursing care at Recovery International Co., Ltd. Elith Inc. has published a case study introducing its AI deployment work with Recovery International Co., Ltd. Recovery International operates in the home-visit nursing field and is proactive in addressing route-planning problems using AI. This case study clearly summarizes how Elith approached the operational challenges the company faced and what results were achieved. This article covers the challenges at the time of adoption and
～Bringing AI closer to everyday life, with video tips for both work and daily life～ Elith Inc. (Representative Director: Koki Inoue), a provider of AI solutions, has launched an official TikTok account (@elith892) to clearly communicate the latest AI technologies and useful ways to use them. On this account, we will share short videos on trending generative AI tools and practical know-how, useful not only in business settings but also in daily life. CEO Inoue himself even appears in the videos, guiding viewers through the appeal of AI from an expert's perspective! We plan to continuously deliver content that is fun to learn from, even for those who think "AI seems difficult" or "I don't know how to use it." Please follow us and experience a daily life with “usable AI”