AI News, FRONTEO (TSE: 2158), Asia's Leading Legaltech Innovator ... artificial intelligence

06 | February | 2019 | Bamboo Innovator

Through our cross-sector and in-depth fundamental research process, the Fund aim to provide access to companies whom we believe are run by high-integrity, honorable and far-sighted owner-operators with a higher sense of purpose in solving High-Value Problems for their target customers and society, and have unique, scalable and wide-moat business models with sustainable competitive advantages and innovative products, services, and processes to create, expand and service their total addressable market, including the resiliency and balance sheet strength to maintain or grow profitability, pricing power and market positions through up/downturns in the economy.

Once we have gotten our ticker code and approval from the regulator for our UCITS fund, interested long-term partners can simply buy or sell units in the fund, just like the buying or selling of listed shares, by approaching any bank of your choice with the ticker code and wire the funds into an independent custodian bank. As a strictly regulated UCITS fund, there is daily liquidity to buy or sell, and we have zero exit fees.

FRONTEO (TSE: 2158), Asia’s Legaltech Innovator Powered by Artificial Intelligence Expanding AI Solution Into Healthcare/BI — H.E.R.O. Innovators Insights from CEO Masahiro Morimoto

“The reason I set up FRONTEO in Aug 2003 is rooted in the mission to solve the problem of Asian companies fighting litigation measures globally, especially in the litigious US society where they are incurring huge costs.

With the desire to protect companies in Asia including Japan, and as a result of thoroughly pursuing technology in a demanding environment where high precision is needed in a short period of time for international litigation support and for fraud investigation, we have created a unique proprietary artificial intelligence engine called KIBIT.

Guided by the mission to illuminate the darkness lurking in data by securely identifying the risks and opportunities buried in records by utilizing technology, FRONTEO (formerly called UBIC) was founded in Aug 2003 by Masahiro Morimoto as a legaltech company supporting eDiscovery (electronic evidence disclosure) and digital forensic services for clients involved in US and international cross-border litigation, corporate investigations and intellectual property disputes.

FRONTEO’s KIBIT series had acquired the number one market share of 26.8% based on sales value in the domestic natural language analysis market for 2016 and 2017 in the independent ITR survey announced on 11 Dec 2018, which is nearly doubled that of the 13.9% share held by the second place vendor.

KIBIT uses the proprietary and unique “landscaping” AI technology to learn the “tacit knowledge” (experience and intuition) of human experts and, unlike all other AI technologies, KIBIT requires only small volume of data for effective machine learning to produce good quality output and solve customers’ problems.

In its legaltech business, FRONTEO has far superior eDiscovery support tools in which Lit i View is the only platform that supports Asian languages without the garbling problem and the only one that offers end-to-end eDiscovery services that combines a workflow system plus AI analytical tools KIBIT that increases the efficiency and accuracy of the document review process, a critical aspect of eDiscovery that accounts for 70% of the cost.

However, despite its overwhelming technological superiority, FRONTEO has faced great difficulties over the years in trying to break through the established practices of US litigation cases that involve Asian companies, its main market, and grow its market share from the current 3%, as well as achieve a stable targeted operating profit margin of 10%.

Thus, unlike the other AI innovators we have investigated whose solutions were readily and rapidly adopted by their clients to solve high-value problems, FRONTEO was an odd case that stood out because the legal industry was particularly resistant to utilizing AI despite its effectiveness since the hours spent can be billed back to the clients and the attorneys make or influence the decisions on the IT vendor based on “relationships”.

positive inflection point in its legaltech business could be near after the completion of structural business reforms at its US subsidiaries in FY18 helped FRONTEO to win several large Asian (including Japanese) clients in 1H FY2019 and the management comments that “We are finally able to do what we’ve been really wanting to do” and expects a strong 2H pipeline with several more orders.

After 15 years since its founding in 2003, FRONTEO finally established a global legal headquarters in US in April 2018 with global offices under direct control of this HQ to strengthen the overall sales structure and better build relationships with US law offices, strengthened internal controls, and undertook actions to curtail unprofitable projects in the US.

In the face of such restraints, FRONTEO has developed a proprietary machine learning technology of ‘landscaping,’ or an algorithm to process natural language, which deftly identifies the specific features of a small volume of data and makes judgement that can be consistently applicable to a larger volume of unknown data, functions in a manner akin to landscaping, or visualizing a complete scene from a narrow field of view.

Although in litigation, e-discovery is the most expensive process, costs can be cut drastically with our AI.” “FRONTEO’s KIBIT series had acquired the number one market share of 26.8% based on sales value in the domestic natural language analysis market for 2016 and 2017 in the independent ITR survey announced on 11 Dec 2018, which is nearly doubled that of the 13.9% share held by the second place vendor.

The e-discovery support business that FRONTEO has been working on since its establishment in 2003 is equipped with the language analysis technology KIBIT series that was developed independently for digital forensics and used in the Legaltech business from 2012, We are developing AI solution business for outside customer companies from the year 2014.

FRONTEO was able to acquire and maintain the position of a leading domestic AI language analysis company serving customers in a wide range of industries including manufacturing industries and financial institutions and used in various business scenes such as customer support, technology development, product development and marketing.

We have expanded the number of companies introducing applications from international litigation support to data analysis in general corporations exceeding 100 companies as at June 2018, a yoy growth by 1.5-fold.” When asked about the origin and meaning of the word KIBIT and how the “landscaping” AI technology works, CEO Morimoto explains: “Learning ‘tacit knowledge’ is a special feature of KIBIT.

But, in behavioral informatics, we are able to predict the future, and we do so by basing our analytics on human cognition and by generating patterns of human and social behavior.” CEO Morimoto went on to elaborate on the applications of KIBIT in financial institutions, its development to process and analyze beyond natural language text data to images and voices, and the launch of the second-generation KIBIT G2 in Nov 2018: “Let’s take financial institutions, for example.

Financial institutions also must check several items, such as the appropriateness of transactions in proportion to customers’ willingness to purchase, knowledge, investment experience, allocation ratios and other factors, inadequacies in the documents offered to customers, consent by customers’ family members, and any unreasonable aspects found in customers in their discussions.

In KIBIT G2, customization of analysis algorithms also became possible depending on customers’ uses, which became applicable to a wider range of applications than before.” CEO Morimoto shared several new developments in KIBIT Patent Explorer 19 and KIBIT Find Answer: “Our patent search and analysis system ‘Patent Explorer 19’ was selected as a finalist (top six) in the Product Innovation Competition at the world’s largest legaltech exhibition ‘Legaltech New York 2019’ to be held from 28–31 Jan 2019.

By simply giving small amount of patent document information as teacher data to the AI engine KIBIT, existing patents in the patent database can be automatically converted into scores and displayed in descending order of similarity, substituting the judgment made by a person in extracting similar patents and speeding the patent search.

The number of companies adopting Patent Explorer 19 has progressed in a wide range of industries from chemistry, materials, machinery to F&B manufacturers and has increased 2.5-fold yoy as at Nov 2018 to exceed 50 In response to the growing demand for patent analysis efficiency improvement.

For example, this can be used in a place where interactive communication such as customer support with a robot occurs.” CEO Morimoto also shares his excitement about the potential of FRONTEO’s second AI engine Concept Encoder: “In addition to text data, we aim to promote the use of medical big data with our new AI solution which analyzes numerical data.

And in the field of digital marketing, we analyze customer’s opinions and apply it to product development and corporate strategy by making use of the tacit knowledge of excellent marketers.” “On 18 Nov 2018, we acquired patent (Patent №6346367) for our second AI engine ‘Concept Encoder’, a epoch-making new search technology dedicated to healthcare to speed up discovery of new drug candidate compounds.

Utilizing a different approach from conventional search or simulation technologies, a major feature of this new search technology is that researchers engaged in the development of novel drugs at drug manufacturers can now use natural language processing technology to search for and investigate the relationship between their hypothesis and the contents of the literature in the public database information.

This function is one of the big strengths of Concept Encoder, which has never been before, which will speed up the discovery of new drug candidate compounds.” “In the R&D department of pharmaceutical companies, updating information on a daily basis by checking the research paper information and the public database is an essential task for quickly finding new drug candidates.

It is important to be able to substitute virtual means such as in-silico analysis or to be able to pinpoint target information from public information.” “The analysis also features ‘light processing’ that eliminates the need for large expensive equipment such as supercomputers and large-scale server groups.

Methods for simulating the action of drug candidate compounds using AI with public database information have been actively conducted and several domestic companies have already achieved R&D efficiency.”“This technology is a practical and convenient technology that can be used not only for drug candidate search but also for analyzing various networks in the future.

Concept Encoder is also capable of co-analyzing with data other than text, and we are conducting co-analysis with numerical data such as gene expression information, vital and various examination values accumulated in the health care area.” While FRONTEO is now profitable and more stable, CEO Morimoto shared that his personal and entrepreneurial journey to start FRONTEO has been full of struggles.

In 2017, we replaced management at our principal subsidiary, FRONTEO USA, as part of restructuring.” “In FY03/18, FRONTEO conducted a sweeping reorganization of the US business, where we eliminated low-profit projects to maximize profit margin and established a unified cross-border sales team to acquire large project wins in Asia by leveraging its proprietary platform Lit i View, which supports multiple East Asian languages, while on the cost front, we managed to reduce costs by scaling back the office responsible for review work, and streamlining the payroll.

Having now assembled the global infrastructure necessary to secure large contracts in Asia, FRONTEO’s plan for FY03/19 onward is to embark on a phase of aggressive expansion, during which we will see to obtain large projects from Asian clients as well as further strengthening the Lit i View platform.” On the growing success of FRONTEO’s AI Solution business for non-legal clients, CEO Morimoto shared his thoughts on growing the business further with the KIBIT Partner Network Program (KPN) and the BizDevOps Lab to promote and support the adoption of AI: “We have been expanding the AI technology cultivated for Lit i View to other applications.

In business improvement, we contribute to increasing our customers’ level of satisfaction by extracting their complaints and signs of dissatisfaction which staff may not be aware of from discussions with customers.” “We have introduced KIBIT to more than 120 domestic and overseas companies (cumulative as of the end of September 2018) such as finance, manufacturers, services, transportation and government agencies, and have been widely supporting business efficiency improvement and social productivity improvement.

Our KPN partners include SCSK (TSE: 9719), TIS Inc (TSE: 3626), System Information Co (TSE: 3677), Mitsui Information (MKI).” “FRONTEO AI BizDevOps Lab, a facility that promotes and supports the adoption of AI, was launched on 16 Nov 2018, to assist customers in overcoming obstacles to AI adoption by creating and optimizing training data and conducting analyses on results derived from AI.

We will be able to more quickly and effectively make proposals to customers and give feedback to its technological team thanks to the unification of our consulting, analysis, and application development teams.” While we find FRONTEO’s AI Solution recurring revenue business to be promising, the legaltech project-based business is troubling.

eDiscovery also enables a verdict based on objective facts and promotes settlements based on the adversarial system (where two parties present their case before an impartial third party), the foundation of the legal system used in the US.” “Discovery is not used in the same way in Japan or other parts of East Asia, so many Asian companies are unfamiliar with the process.

It is a kind of recurrent revenue for us.” “FRONTEO’s proprietary platform Lit i View is the only solution that offers companies end-to-end eDiscovery services that combines a workflow system plus AI analytical tools KIBIT that increases the efficiency and accuracy of the document review process, a critical aspect of eDiscovery that accounts for 70% of the cost.

As an example, we have developed a review management dashboard where clients can have up-to-the-minute metrics on review progress, which includes a heat map within Lit i View that provides a detailed snapshot of reviewer progress and productivity and enables course corrections as needed.” “Conventional eDiscovery tools developed in English-speaking countries cannot accurately process legal documents written in Asian languages or multi-byte characters in Chinese, Japanese or Korean (CJK), without experiencing problems such as garbling.

Asian companies, which thus are at a disadvantage in terms of the e-discovery process, have found that our own e-discovery reviewing tool Lit i View that use our proprietary KIBIT AI Predictive Coding provides an effective solution to their problems and improve the quality of reviews, which can account for up to 70% of eDiscovery costs.” “We help our CJK clients understand how we’re going to organize a complex case that potentially includes multiple languages.

In eDiscovery, there has been no significant change in litigation costs or customers’ litigation budgets, yet there has been a steep rise in the volume of data involved (higher customer costs) and greater pressure to remain cost competitive with every passing year FRONTEO aims to take advantage of this situation improving upon Lit I View.

We think that our competitors will face difficulty adapting to the changing environment, since their services rely on third-party tools developed by other companies.” “Today, risk cases such as legal risks such as lawsuits and compliance violations at home and abroad, internal control problems arising through accounting and auditing, cyber security measures such as information leakage from outside invasion, etc.

In projects where third-party committees are established in recent years, the amount of data to be surveyed and investigated is large and response to report in a short period is required.” “In response to these circumstances, on 21 Jan 2019, FRONTEO announced our Legaltech Solution Lab which integrates DOC (Discovery Operation Center) and Digital Forensic (Electronic Identification), the two facilities of eDiscovery in International Lawsuit, with FOC (Forensic Operation Center) which conducts fraud investigation.

We will utilize practical know-how, such as forensics, discovery, cyber security measures, third party committee surveys, etc., to conduct large-scale data surveys and prompt response to risk deals with our proprietary AI engine KIBIT and the Lit I View series and the research capabilities we have cultivated over many years.

We will make maximum use of our know-how in legal affairs and internal control and provide early detection and resolution of problems to a wide range of companies.” “In addition to increasing the scale of data and shortening the response time, Legaltech Solution Lab also responds to changes in the information recording environment, survey method, fraudulent means and signature of information surrounding companies as follows: “(1) In discovery, records that were once in servers and PCs are also stored in the cloud and mobile terminals, so the importance of cutting-edge digital forensic research techniques such as combining data analysis and log behavior analysis is increased;

(4) When cybersecurity incidents occur, movement to prevent damage confirmation and secondary expansion, such as identifying spill data and intrusion routes, and taking the next measures urgently is required.” On the problems faced by FRONTEO’s legaltech business and how they are solving them, CEO Morimoto said: “When a lawsuit is filed in the US, Asian companies have a strong tendency to leave the selection of lawyers and other service providers to their local subsidiary.

This is because lawyers tend to be more loyal to contract partners and lawyers themselves are salespeople who will be dismissed by their firm if they do not win clients.” “In the past, the company has laid foundations for future development by establishing a US subsidiary (2007), listing itself on the NASDAQ (2013), acquiring US eDiscovery providers (2014, 2015) and improving our visibility and law office network in the US (through M&A).

Under the direction of our global legal HQ, we have integrated our Asian and US sales operations and are working to build and strengthen relationships through initiative such as roadshows that visit US law offices and local subsidiaries in Japan and the rest of Asia.” “These roadshows serve to show how Lit i View can help make the eDiscovery business more efficient and to secure information concerning litigation support projects for Asian companies at an early stage and contracts for the eDiscovery business.

In order to protect Asian companies from these disadvantages, FRONTEO has been developing and improving local discovery services (eDiscovery support work in the home country), Asian-language compatibility technology and AI technology to efficiently identify evidence.” “Current trend where eDiscovery data continues to increase year after year is favorable for FRONTEO, whose strength is AI technology.

These cross-border sales efforts have already started to demonstrate positive outcomes in winning large projects from companies in Asia.” “After successfully completing the operational reform during the fiscal year ended March 31, 2018, FRONTEO is now able to start its global sales development aggressively.

As for Lit i View, which is the service platform supporting our litigation support service since 2010, we are currently developing the next generation to speed up scalability improvements and to increase flexibility in the user experience of our software while maintaining its all-in-one feature advantage, which is incomparable to other vendors’ technology.

We are aiming to acquire an overwhelming share in the global market with our unique AI solutions that other competitors cannot offer.” CTO Takeda Hideki also shared how he was attracted to join FRONTEO because of the philosophy and mission of FRONTEO to bring about the realization of fairness and to protect companies in lawsuits, as well as his views on the legaltech business, including a new pricing model to charge by the documents reviewed instead of by the hours so as to better estimate the cost beforehand which is more desirable and fairer for the end client: “I joined FRONTEO in 2009 to lead the R&D of AI technology and FRONTEO introduced to the world the first AI forensic tool in automatic email auditing using machine learning.

In order to further evolve this and extract information more quickly and with high accuracy, we developed FRONTEO’s proprietary automatic document analysis technology ‘Predictive Coding’ that applied natural language processing, text mining, machine learning and other techniques.” “FRONTEO further developed Predictive Coding into the artificial intelligence engine KIBIT which analyzes not only conventional clustering and trend analysis but also the qualitative approach based on the outcome of the case database and behavioral science, using the tacit knowledge of human beings, making it possible to predict human behavior and community visualization.

We can create technologies with our own hands that make society better and feel the responsiveness of the contribution to society and customers, and the challenge as a business at the same time.” “There are several types of legaltech business markets, but the big one is an area related to international lawsuits centered on the United States of e discovery, which is growing at 15% per year.

Service vendors offering e-discovery outsourcing services will be forced to face a tough battle in the future, and it is expected that the consolidation of US service vendors seen in recent years will continue.” “On the other hand, technology vendors that provide software to the industry never deliver products directly to end users;

Looking at the trends of technology vendors in the legaltech industry in the past 10 years, even though vendors whose company size have grown to a reasonable scale and are acquired by major IT companies, the new developments are stagnant after acquisition and they disappear as players This situation is continuously seen.

Regarding FRONTEO as a technology vendor, in addition to the strength of being able to process languages ​​other than English correctly and with the natural language analysis AI engine KIBIT as the core, FRONTEO has focused on areas other than e-discovery of evidence in litigation, analyzing analyze various kinds of text data such as patent information analysis.

In addition, Lit I View, our service platform from 2010 which has supported the FRONTEO litigation support service, has been improving scalability and maintaining all-in-one goodness not seen by other technology vendors.” “In advancing the development of the second-generation Lit i View and AI Review software, it is extremely important that we have the face of both service vendors and technology vendors.

What I’d like to realize is to develop AI as a daily weapon of intellectual professional to bring about the philosophy of realizing fairness.” CTO Hideki went on to elaborate patiently about KIBIT’s “landscaping” AI technology and the various innovative applications: “Many artificial intelligence in the world aims to become smarter by analyzing big data, but there is a need to ‘teach’ the artificial intelligence using a lot of data so that it can actually make effective judgment at the business site.

By extracting human tacit knowledge as a small scenery (a small number of teacher data), KIBIT can form the scenery of the unknown world and also connect to the outside as a single landscape view to learn human abstract judgment and construct knowledge of scale.

Based on ‘behavioral information science’, we embody the data analysis know-how that we experienced as ‘behavior model’ making full use of sociology, criminal science and psychology, accumulate its know-how in a reusable manner, and realize highly accurate data analysis and unstructured data (such as electronic medical charts) analysis in a wide range of fields.” “There are four main business areas in which KIBIT is used.

In response to this issue, hospitals have taken actions to avoid the risk of falls, such as creating and assessing risk assessment sheets, substituting with safe footwear, controlling the amount of sleeping pills, measures which reduced the fall rate by 15%.

However, since the above-mentioned efforts are not the original work of the nurse, the burden of the nurse has increased and a new problem has been created.” “KIBIT analyzes the unstructured data in electronic medical record to predict the patient falling down and the accuracy is improved by about 50% compared with the conventional method such as Morse Fall Scale and STRATIFY.

What most people don’t realize is the majority of the data in electronic medical record that’s valuable is unstructured data, the actual, free-form text that’s used by the doctors and nurses to really understand the whole of a patient which is a treasure trove of tacit knowledge.

The Cancer Research Institute utilizes technologies, such as Clinical Sequence to comprehensively analyze genetic abnormalities in cancer patients and Liquid Biopsy to examine gene abnormality by blood test, and research methods for integrating and analyzing patient diagnostic data such as genome analysis information.” “The second is an informed consent support system.

Unlike general artificial intelligence that learns by inputting an enormous amount of data regardless of whether it is useful or not, the technical advantage of KIBIT is that it produces good quality output by only learning a small amount of good quality data based on the tact knowledge made by learning expert judgment and sense.

By objectively analyzing such tacit knowledge with the artificial intelligence and scoring a new record by severity of chronic pain, it becomes a system to utilize for accurate diagnosis and transition to appropriate treatment leading to pain improvement as soon as possible.” “In collaboration with second-hand car sales giant IDOM/Gulliver (TSE: 7599), we use KIBIT for Gulliver’s online customer service ‘Car Connect’ and provide automatic customer service support to users.

You can find and use the treasure contained in big data, and by using it well, we can truly respond to what the customers are requesting and to improve marketing measures.” “Mitsubishi Heavy Industries makes use of KIBIT’s Knowledge Probe to improve the efficiency of investigation report preparation work aimed at grasping management judgment and industry trends.”“Prior to using KIBIT, the company collects information posted in various news media, analyzes its contents, and summarizes it as a report for management.

The high degree of accuracy of KIBIT demonstrates the enormous power in IP document investigation work which is indispensable when developing new products and in reducing work man-hour to about one fifth of conventional work.” “In collaboration with Toyota Technical Development supporting the Toyota Group’s intellectual property strategy, we developed the patent research and analysis system ‘Patent Explorer’.

From the contents of the patent, we have expanded the range of applications by ourselves using new methods of use, such as technology trend survey investigating latest technical information.” “KIBIT AI extracts the features and concepts related to the contents of the description of the patent rights to be invalidated, and then assigns a score indicating the respective level of proof to each piece within the enormous quantities of patent publications.

Following the major trend toward open innovation and AI, we are confident that Patent Explorer19 will continue to expand and be implemented widely and globally.” CEO Morimoto summed up on FRONTEO’s philosophy to bring about the realization of “Bright Value” and motto of “Enthusiasm, Persistence and Impression” as his driving force personally and for everyone at FRONTEO: “What we mean by ‘Bright Value’ is to securely identify risks and opportunities buried in records.

05 | February | 2019 | Bamboo Innovator

Through our cross-sector and in-depth fundamental research process, the Fund aim to provide access to companies whom we believe are run by high-integrity, honorable and far-sighted owner-operators with a higher sense of purpose in solving High-Value Problems for their target customers and society, and have unique, scalable and wide-moat business models with sustainable competitive advantages and innovative products, services, and processes to create, expand and service their total addressable market, including the resiliency and balance sheet strength to maintain or grow profitability, pricing power and market positions through up/downturns in the economy.

Once we have gotten our ticker code and approval from the regulator for our UCITS fund, interested long-term partners can simply buy or sell units in the fund, just like the buying or selling of listed shares, by approaching any bank of your choice with the ticker code and wire the funds into an independent custodian bank. As a strictly regulated UCITS fund, there is daily liquidity to buy or sell, and we have zero exit fees.