“Thirty years ago a person who was able to operate a personal computer was considered a hero, but everyone became able to use it like a tool thanks to Bill Gates. After ten years, the Internet became popular and It became another tool which require us to be able to master it. Smartphone appeared ten years later. Initially it was a toy-like device, but due to the increasing sophistication and popularization of app functions, everything including digital wallets entered into smartphone, creating dramatic changes in life. We believe that the spread of RPA and AI will fundamentally change and redefine the structure of many businesses and raise the productivity of white collar workers dramatically in the coming decade. RPA will be a ‘stationery’ which is easily used by anyone without any special knowledge and that every white collar worker will use every day. Everyone can design and create a RPA software robot that matches their business. Then, even if there is a change in the work flow and business rules change from tomorrow, people in the field can change on the spot and respond immediately. The history of productivity revolution using technology is a series of such things when they make it to the level that you can use it like that. RPA will become the reading and writing of the future era. Personal computer, internet, smartphone and RPA will be basic business skills and RPA will be a powerful weapon for many companies. The future of the company and the Japanese economy will change depending on how you use the time of the person who created the RPA software robot. You can devote your free time to designing more efficient business processes, or you can use it for developing new businesses and products with higher added value. What is important is getting the degree of freedom about how to use the time created by productivity improvement,” commented Tomomichi Takahashi, founder and CEO of RPA Holdings (TSE: 6572) (“RPAHD”).
RPAHD is Japan’s #1 SaaS (Software-as-a-Service) innovator in AI x RPA (robotic process automation) with an overwhelming market share of over 80% (by both sales value and number of clients) while the penetration rate is still low at less than 5%, indicating a large total addressable market and a long runway for growth. RPAHD operates two key businesses with a recurring-revenue subscription business model:
(1) “Robot Outsourcing” (40.3% of sales, 69.8% of operating profit) with its flagship cloud product BizRobo! automating manual repetitive high-volume workflow-intensive back-office/shared-services tasks where increasing the speed and accuracy of a process will lead to outsized benefits, carrying out transaction monitoring to continuously check if transactions are still in compliance, learning fraud patterns to prevent fraud etc, serving over 1,000 clients with recurring monthly subscription revenue that include Nippon Life, MUFJ Bank, ORIX Group, Unilever Japan as of January 2019, up exponentially by over 10-fold from 91 companies in Feb 2017 and growing rapidly from 250 in November 2017 and 440 in September 2018.
(2) “Robot Transformation” (56.8% of sales, 36.8% of OP) with its performance-based marketing service PRESCO using RPA to automate the online B2B advertising affiliate work and disrupt the conventional ad agency business, serving clients that include M3 (TSE: 2413) and Aiful Corp (TSE: 8515), one of Japan’s largest consumer finance companies.
Two overlooked major strategic business partnership breakthroughs are made with global software giants Microsoft and SAP:
(1) Microsoft Japan has collaborated with RPAHD to jointly develop Robot-as-a-Service on Microsoft Azure and offer the solution since Aug 2018 not only to quickly and flexibility create and introduce a cloud environment necessary for RPA but also to develop new functions using AI.
(2) On 10 Dec 2018, German ERP giant SAP announced that it will used RPAHD’s BizRobo tool to automate ERP input to SAP S/4 Hana cloud and SAP ERP for the SMEs customer base. When introducing SAP ERP, not only does it takes a lot of manpower and time to input the initial data, but it is also required to maintain the freshness, precision and granularity of information in the ERP continuously after the operation is started. SAP commented that utilizing BizRobo will speed up the completion of introduction of its SAP solution from 6 months to about 9 weeks by reducing the work change and operation burden during implementation.
An RPA system “observes” the user perform a task in the application’s user interface and the various decision points involved in accomplishing that task, and then perform the automation by replicating the process in repeating those tasks directly in the in the user interface. Unlike an Excel Macro which performs a single task, AI RPA can execute a series of complex tasks. RPA is non-invasive with no changes or additional coded interfaces in existing IT applications (sits as on-top layer) and can readily run on any in-house platform or cloud-based system, and can interact with multiple applications at once. Compliance risks are minimal because every action executed by an RPA bot is logged and thus auditable. RPA requires no knowledge of programming or code and people can start automating processes with RPA tools with only a few days of training. RPA requires few changes to the underlying back-end systems. Such a lightweight approach to automation lowers the threshold of processes worth automating. RPA integrating cognitive AI elements such as natural language analysis, image recognition and machine learning can handle processes with unstructured data, identify patterns, understand the context of text-heavy content, make rule-based decisions that don’t have to be pre-programmed (machine learning), and offer customers suggestions (cognitive agents). While a macro is fixed and linear, a RPA is dynamic — it can “learn” and respond to stimuli (events, occurrences etc), accumulating knowledge of procedures over time, thereby getting “smarter”. It is estimated that up to 45% of the activities companies pay people to perform can be automated through RPA.
With the compellingness and circumspection exponential edge in delivering automated services for its growing clients, RPAHD achieved a 244% absolute increase in sales and 446% in operating profit in the recent three years achieving growing profitability with estimated FY2019/02 operating margin of 10% (3Q FY2019/02: 9.8%), ROE (= EBIT/ Equity) of around 28–30.5% and ROA of 12–13.9%, propelling a 325% increase in market value since its listing on 27 Mar 2018 at the IPO price of 3,570 yen per share (split-adjusted price of 714 yen) to US$739m. On 15 Jan 2019, RPAHD announced its 3Q FY2019 results (Mar — Nov 2018) in which 9M sales jumped 99.8% to 5.83bn yen and operating profit increased 61.4% to 557m yen. In the robot outsourcing business, BizRobo! sales grew 113.0% yoy to 2,350m yen while operating profit increased 52.1% yoy to 389m yen. In the robot transformation business, PRESCO sales grew 122.6% yoy to 3,114, yen and segment profit increased 112.1% to 205m yen due to expansion in the use of online advertisers driven by the continuous demand in the field of medical personnel for nurses and pharmacists which PRESCO specialized in. Balance sheet is relatively healthy with net cash of 1.156bn yen (gross cash 3.267bn yen, gross debt 2.111bn yen), or 1.3% of market value. Management estimates FY2019 sales to increase 2.2-fold to 9.1bn yen and operating profit to rise 94.8% to 906m yen. Softbank is a strategic shareholder with a 4.45% direct stake, while SBI Investment & SBI Securities own a further 6.93%. RPAHD is listed on TSE Mothers in March 2018 and is in the process of upgrading to the TSE 1st section. In Feb 2018, RPAHD has adopted a stock option system of 859,000 shares, equivalent to 16.6% of outstanding shares, to incentivize management and employees.
Global RPA companies include the loss-making London AIM-listed RPA pioneer Blue Prism Group (LON: PRSM) which has a market value of US$1.2bn with sales of US$72.5m vs the profitable RPAHD’s market value of US$752m with sales of US$82.8m. In August 2018, Blue Prism announced that it reached the milestone of more than 1,000 client companies using its software. Prominent RPA startups which are still loss-making include UiPath, which raised US$265m in November 2018 at a US$3bn valuation with over US$150m in sales and 2,100 enterprise customers that include BMW, Autodesk, Huawei; and Automation Anywhere, which raised US$300m in November 2018 at US$2.6bn valuation with 1,400 customers. RPAHD is possibly the only profitable pure-play RPA SaaS company globally thus far.
While RPAHD is profitable and positive free cashflow generative now, CEO Takahashi shared that his personal and entrepreneurial journey to start RPAHD has been full of struggles. CEO Takahashi shared reflectively the inspiring story: “I was born in Fukuyama city, Hiroshima prefecture in 1970. After graduating from the School of Economics at Hitotsubashi University in 1993, I joined Accenture which was a pioneer in starting management consulting to introduce computers to businesses such as GE for the first time in the world in 1953. When I joined the company, I was assigned to the system development project in a program focused mainly on CBT (Computer Based Testing). In retrospect, I was able to feel the possibility of the digital information revolution during my three years with Andersen Consulting, as I looked back to the use of computer by businesses at the dawn of the digital information revolution. Furthermore, because of the many financial-related projects, I gained knowledge of financial accounting and capital markets. Also, since it was necessary to produce the output to the client within the deadline, even if they are in unknown industries and unknown fields, I was trained in the speed of learning and outputting. I think that it became a great asset in my career in learning to be an expert in new businesses after graduating from Andersen.”
“In 1995, I read the bestseller Being Digital written by Nicholas Negroponte, who served as the first Director of the Media Lab at the Massachusetts Institute of Technology in 1995, and learn about the coming digital information revolution. At the time, Softbank, which had less than 100 billion yen in sales and was handling personal computer magazines and software distribution and was just listed on the JASDAQ, launched the current Sky Perfect TV jointly with media king Rupert Murdoch to enter into the digital multi-channel broadband broadcasting business. Having been inspired by Professor Negroponte’s book and hearing about the story of the launching of Sky Perfect, I immediately wrote a resume and was I was lucky to be hired as a startup member. I was about 26 years old. Later in joined the main body of Softbank and was involved in the launch of Softbank Finance, now SBI Holdings.”
“While most of my classmates aim for large companies and are excited by talks about lifelong wage, I felt a very large value gap in conversation with my classmates. Because my father is engaged in the printing industry and many of my relatives also owned small businesses themselves, I grew up from early childhood watching how we talk about our business, how we will challenge and expand our business. I really enjoyed talking about how we shared the results with employees in the form of salary or bonus. It was in April 2000 that I started a company with four people, including Nobuyuki Osumi, CEO of our wholly-owned subsidiary RPA Technologies.”
“However, this was just after the burst of the IT bubble and the capital market crashed completely and it was a very harsh environment in securing financing funds for venture companies. Meanwhile, I grasped the opportunity that large companies began to tackle new Internet-related business, but since they lacked talented people and know-how, we were engaged in business consultation work specializing in new business using the Internet and I was able to grow the company. We were dedicated to this project until 2008, with annual sales at 700 to 800 million yen, and employed 30 people. However, 40% of the ongoing projects stopped when the Lehman shock broke out in 2008 and we faced a crisis of survival. Fortunately, we had special talent and internal reserves necessary for structural transformation. I made a decision to change the direction of the company.”
“In 2008, before the concept of robotic procession automation (RPA) was proposed, we were introduced by a major consumer electronics maker, who was our client, to Denmark’s Kapow Technologies who possessed the core technology of RPA. was acquired by Kofax in July 2013 who in turn was acquired by private equity firm Thomas Bravo in July 2017. Inspired by what we see that ‘this is a white-collar version of a robot!’, we started working on developing BizRobo! to automate complicated and troublesome tasks and processes in day-to-day operations, such as accounting, auditing, personnel affairs, legal affairs. The BizRobo was conceived as a white-collar robot in the service productivity revolution, in the same way a factory automation robot at the factory production site works. In the first place, the administrative department is uneven in the amount of work depending on the day. For example, in the case of accounting, RPA software robot can automatically handle all the journalizing. The important thing is how to utilize the free time thereby to tackle various things. Those who wish to do more high-value work and those who want to spend more time with their families can choose to do so.”
“Although we struggled hard in sales initially, our delivery of BizRobo to Nippon Life, a major life insurance company in 2013 was a turning point. Their back office work which had been handled by 80 people was reduced to 13 people by introducing 15 RPA software robots (5 for data inputs and 10 for data checks). Starting with the life insurance industry, our introduction of BizRobo is progressing into a variety of businesses in multiple industries, such as banks, communication, aviation, travel, media, property insurance, retail, distribution, apparel.”
“There were also publicized media reports of successful use of RPA to improve efficiency in MUFJ Bank, Mega Bank, ORIX Group and Softbank. However, we were still unable to really expand our business until from around 2015 when we break out. Although we struggled hard in sales initially, our delivery of BizRobo to Nippon Life, a major life insurance company was a turning point. Their back office work in checking customer information which had been handled by 80 people was reduced to 13 people by introducing 15 RPA software robots (5 for data inputs and 10 for data checks), achieving 84% in cost savings. Starting with the life insurance industry, our introduction of BizRobo into a variety of businesses in multiple industries, such as banks, communication, aviation, travel, media, property insurance, retail, distribution, apparel, is progressing, including front-office project works and publicized media reports of successful use of RPA to improve efficiency in MUFJ Bank, Mega Bank, ORIX Group and Softbank. However, we were still unable to really expand our business until from around 2015 when we break out.”
“Professor Leslie Wilcocks and Mary Lacity of London School of Economics coined the term ‘RPA’ in 2015, triggering attention and interest amongst research companies and consulting firms that RPA is the next-generation management skill to acquire. Furthermore, the AI boom accelerated the spread of RPA. The number of client companies rapidly increased from 91 companies in Feb 2015 to 250 in Nov 2015 and has now exceeded 1,000 in Jan 2019. In 2014, there were almost no companies doing RPA services in Japan. However, after 2017, many companies entered the RPA market one after another because of the strong demand. Our domestic market share exceeded 80% and we have introduced RPA in more than 34 internal operations and to over 1,000 companies to greatly improve productivity.”
CEO Takahashi also shares his perspective on the unique RPA situation in Japan: “The office work in Japan is individualized for each company, and there are many customized work. So if you try to automate your work using office automation (OA) or business process re-engineering (BPR), it will cost more. Because of this problem, Japan’s OA has not progressed so far. To solve this problem, RPA aims to solve with the idea that business flow is analogue and office work can be automated using RPA software robot without changing the flow. As an example, let’s consider the case of collecting paper receipts and posting in the accounting department of the company. Until now, information such as the amount stated on the receipt, company name, date of payment, and so on, was input manually to each cell of Excel or ERP accounting system, and a huge amount of work was required. According to the idea of OA conversion in the past, it is said to introduce a system that can exchange payment information online without issuing a paper invoice in the first place, but in Japan such thought did not spread easily. On the other hand, RPA says that it is inevitable to receive a paper receipt itself, and instead it scans the receipt and turns it into OCR, and automates the process of filling each numerical value in the target cell. The idea of such RPA is drawing increasing attention.”
“It is often said that Japanese people have very good abilities when viewed on an individual level, such as mathematical literacy which ranks high in the world. However, according to the OECD survey, Japan’s labor productivity is the lowest in G7. This situation has not changed for a long time. When an excellent individual enters into the organization, he or she cannot demonstrate their ability sooner. The cause is that there are many cumbersome tasks in the company, such as stamping on the form and circulation of the approval note, and individuals are being forced to handle such miscellaneous work, leaving less time to creative work, so productivity does not rise at all. Also, when you get such extra work tasks, those with more free time would slowly go through their existing work and began to pretend to be working hard. Japan’s labor productivity can be greatly improved if this miscellaneous affair can be automated. RPA is effective as a breakthrough.”
“RPA is most effective in environments where there are multiple systems that are not unified in the company. A representative example is the financial industry. The core system of financial institutions is independent from the Internet due to security problems, and it prevents intrusion from the outside. Therefore, in order to transfer the information on the mission-critical system to another system, it was necessary for human beings to visually check and repair them. However, using RPA makes it possible to automatically pick up numbers on the screen and transfer data even if there is no connection between the systems. BPM (business process management) tool that enables such management and reconstruction of the business itself is rapidly advancing cooperation with RPA, and RPA will become a part of BPM.”
Noting that RPA is often misunderstood as a simple macro tool, CEO Takahashi explained further on the difference, as well as RPAHD’s software architecture and pricing model: “The difference between Excel Macro and RPA is that In order to assemble a macro, you will need to gather personnel with expertise on macros and do programming work. So with this method, the company has to pay enormous cost to achieve automation and only for a single task. In the case of RPA, on the other hand, by simply pressing the start button, we will automatically record (log) the flow of series of office work actually done by humans and remember it. When you select one step in the flow, there is also a feature that the screen scene is displayed, so it’s easy to extend, change and modify. RPA’s groundbreaking point is that you can teach the work of human beings directly without relying on programming.”
“A small agent on the desktop communicates with robots on a central server — a different architecture for RPA vendors in that no robot needs to be virtualized or installed on the desktop. The benefit of this is that robots are in one virtual machine that is centrally managed, and there is no need to update individual robots or to push them out to individual machines.”
“As US research firm Gartner pointed out when they selected RPAHD in their Cool Vendors in Business & IT Services 2017 report, BizRobo allows RPA to be used for quick POCs and short-term projects, coupled with scalability over the long term, based on consumption-based pricing. For example, a buyer can pay $200 to use a robot for a month. Another option is 5 cents per line of code, or from $6,000 to $10,000 to rent a robot in a business unit. In 2017, several components of Blue Prism’s RPA and Nice’s RPA were added to BizRobo as an option for quick POCs and short-term projects. BizRobo is offered with contract options such as trial, pay-as-you-go and rental. Customers can select the best contract type to suit their purpose. By providing these flexible contracting options, BizRobo can solve a problem, such that the license terms of general RPA tend to be inconvenient and expensive for short-term projects. BizRobo enables clients to try RPA, use RPA at specific events for a short period, and test a hypothesis on new business using RPA. It also allows SMBs to benefit from RPA, even though RPA has mainly been used by large enterprises.”
RPAHD is also leading the RPA revolution in Japan through education. CEO Takahashi added: “We have established the Japan RPA Association in July 2016 to contribute to the spread of digital labor from a neutral standpoint. Our association is doing RPA education and also established the ‘administrative and academia subcommittee’ to support RPA education for students at universities, high schools, technical colleges, vocational schools, etc since younger people have more adaptability to new technologies. Also, if you take a two-day course, anyone will be able to do RPA.”
“In collaboration with the Japan CFO Association, we organized the AI X RPA Subcommittee to nurture the ‘next-generation accounting professional’. We have a business alliance with Business Brain Showa-Ota’s (TSE: 9658) Management Accounting Information System to set up a BPO service model of RPA. In addition, we operate the content media ‘RPA BANK’ in April 2017 which specialized in providing information, news, case examples on artificial intelligence and RPA, success stories and failure cases, as well as holding regular seminars and events for members. As RPA spreads from now on, and successful cases increase, more companies will be able to think that ‘This may be possible for themselves.’”
Intrigued by the wide and diverse applications of RPA, we asked CEO Takahashi to share more cases of both success stories and failures. CEO Takahashi patiently elaborated: “Actually, the business areas where RPA plays an active role are diverse. They include ‘Business process’ such as contract management clerical work and customer management affairs; ‘information survey’ such as price survey, patent search and fraud detection; ‘EC business agency’ such as product registration, inventory management, order acceptance; and ‘marketing’ such as recommendation and SNS contribution. When introducing RPA, we think that there are three patterns: (1) ‘Routine response type’ to automate routine work; (2) ‘Idea execution type’ to automate complicated routine tasks based on their own hypotheses and intuition, highly experienced people, ad-hoc business and exception correspondence by deep learning; (3) ‘innovation type’ that collaborates with various companies, local governments, organizations, etc. to automate and create new businesses and jobs, including professional digital labor business and regional shared service.”
“For example, in issuing invoices. it took about 3 days to process 400 sheets with human hands. When we leave the task to the RPA software robot, the operation finished in 35 seconds per sheet. Moreover, the company said that sales have increased three to four times over the past three years, and the number of invoices to be handled proportionately increased sharply. Nevertheless, thanks to the RPA software robot, the work time is still 35 seconds per sheet. In a food company, what used to take them two days to compile inspection of food ingredients now takes only 8 seconds. In the case of ORIX Group’s Okinawa Business Center where 800 staff members are contracted to handle various administrative processes for its 12 group companies, the amount of administrative processing fluctuates according to the period in some departments. They were unable to deal with rapidly increasing administrative processing and were suffering from high staff turnover or confusion which leads to mistakes. However, since the person-in-charge of the business department is not familiar with IT, he wants to avoid the introduction of a complicated system and succeeded by leaving the work to RPA.”
“At the Bank of Tokyo-Mitsubishi UFJ, as a result of introducing RPA, they succeeded in reducing work by 8,000 hours per year in 20 kinds of office work. In May 2017, it created a digital strategy unit, and has sped up application of the technology. By the end of fiscal 2017, the firm plans to have about 100 tasks being performed by RPA. And in the six years starting from fiscal 2018, the bank plans to apply RPA to some 2,000 tasks. A dedicated team is examining the bank’s operations, and choosing which tasks can be automated. System developers then create programs that can do the selected tasks efficiently. For the time being, RPA will be employed for work related to checking home loan application documents, interbank transactions, and international remittances, among other functions. Mizuho Financial Group, meanwhile, is also looking to have RPA take over 100 tasks once performed by people by the end of this fiscal year, saving 300,000 work hours annually, equivalent to the load on 150 employees. The group’s Mizuho Bank Ltd. already uses RPA in the process of opening investment trusts, automating tasks that had taken human employees four to five minutes and compressing them to a few dozen seconds.”
“SMFG’s Sumitomo Mitsui Banking Corp. began introducing RPA systems in 2017. In just the first half of the 2017 business year, RPA reduced human tasks by 400,000 hours, equivalent to an entire year’s work for 200 people. The firm is aiming to ramp the system up to save some 3 million human work hours annually by March 2020, equivalent to the annual workload of 1,500 employees. From April 2018, the Sumitomo Mitsui RPA will automatically collect all of a client’s investment and asset data by the morning of any day that client has an appointment at the bank. The system will then send that information package to the banker in charge of the client’s file. The bank has also begun training courses to help promote the use of RPA among its staff.”
“At a major communications company, they face problems in responding quickly to customer calls at their call centers. After receiving a telephone call from the customer, the staff investigates and confirms the content of the customer inquiry and replies accordingly. However, since it takes time for administrative tasks such as inquiries and confirmations for each case, there was a limit to the number of cases that could be handled. There were 30 to 40 screen transitions that occurred in response to an inquiry, and it took about 20 minutes for one process. Because it is necessary to instantly select the optimal one from the branching options for each condition, it was easy for mistakes to occur. Processing speed varies greatly between newcomers and experienced staff, causing a difference in waiting time. Education of new staff also took time, so it was always a shortage of people. RPA greatly improves processing speed in making only transitions to 3 screens or less from 30 to 40 screens. As a result, the time taken for one process was 1 minute, which was sharply reduced compared to 20 minutes before introduction. RPA is responsible for the work that requires accuracy and promptness. By promptly selecting the right conditions under complicated rules, the mistakes also disappear. Because the waiting time has decreased, the customer’s stress is also gone, and the claims were drastically reduced. The work that took 10 staff has become possible with one new staff, which also leads to reduction of personnel and cost.”
“The cost of introducing and maintaining RPA is much lower than hiring one temporary staff and you can expect high cost effectiveness. The era in which each department hires temporary staff is over. Cases are also emerging where the precision and speedy work becomes a great business weapon. In the example of a mortgage company, the contract rate is raised by automatically processing a loan application faster than other companies.”
“PRESCO is a performance-based promotion service that connects advertisers and partner website operators such as M3 (TSE: 2153), with the RPA marketing solution automating the collection, aggregation and reporting of marketing data such as search keywords and search ranking data that supports purchase · estimate · customer acquisition and marketing automation in content marketing, specializing in job seekers in the medical industry such as pharmacists and nurses.”
“Our marketing RPA software robot can substitute collecting information using search engines for EC (ecommerce) companies. For example, if you are a used car sales site, you need to grasp information on quoted prices about ‘How much cars can you sell at present’ by traveling to other companies’ sites before determining the price of products. It is a laborious task, but this can also be automated by RPA.”
“Veteran sales representatives of used car sales can perceive the market trend sensibly while traveling around several sites. However, such an ambiguous ‘intuition’ has a problem in that it is hard to understand other than those who have experience for many years. On the other hand, if you use RPA combined with AI, it will automatically traverse the Internet, gather information and automatically extract the market trend from that data. There are also companies where salespeople regularly check out review sites and blogs and report the topics that they find exciting. They raise the productivity of the sales force by automating and distributing the report regularly to sales offices nationwide.”
“One of our clients who is an apparel company used to check the turnover rate for each item on a daily basis to discover ‘dead muscle’ items that she does not think is likely to sell. Delete this ‘dead muscle’ product from inventory and register the item on a membership site that trades apparel products, and upload the notice of the discounted sales item to SNS. By doing all of this series of work by the RPA software robot, it speeds up inventory disposal of ‘dead muscle’ products, leading to a significant improvement in profit margin.”
“At Unilever Japan, they would like to smoothly guide customers who are interested in each product from visiting their brand sites to the EC (ecommerce) site and to raise the cost effectiveness of their product campaigns by raising the conversion rate from the brand site to the EC site and contribute to sales. They were acquiring information such as EC site inventory, price, return rate, etc. for over 3,000 items. Even if three people continue to work 24 hours a day with manual labor, it will take more than a week, and the work is also complicated by multiple languages. RPA automated the acquisition of information related to their products sold on various online channels and displayed the information in real time. As a result of becoming able to immediately grasp important consumer needs such as price, shipping cost etc, it is possible to grasp the sales trend of their products and make effective strategy. RPA is also incorporated an internal tool for improving and managing OSA (on-shelf availability or percentage of products in store). Usability improved and traffic from the brand site to the EC site rose.”
“RPA is also used as a cyber patrol to check the listings on the EC site for illegal items and remove them. Currently, there is a trend called Open Source Intelligence (OSINT) that analyzes various information by collecting various information from multiple public information sources including Web sites and social media, and is especially useful in detecting illegal or strange information such as whether the uniform of the bank security guard or bank clerk is sold on the net. OSINT is also selected as a defense technology in the United States FBI, CIA, Defense, Navy, Police, British police, etc. Since 2012, we have been producing RPA results in the area of compliance. Since digital labor can perform 24 hours, the needs are great.”
“A typical failure case is a case where consideration is given to ‘introduction of RPA as the goal’ due to instructions from the top management and others. Although we can introduce only the form, it cannot be used immediately to yield results. It is important that the design of the RPA software robot is in-house production in the end, while receiving external support at the beginning. On the other hand, a successful case is one where the effort could be scaled and deployed to other departments and other tasks. It is important to develop skilled workers who can add, modify and delete tasks to the RPA software robot in the subsequent operation. To that end, you will need to positively adopt young talent who can master the RPA.”
Noting the rapid growth of competitors such as Blue Prism, UiPath and Automation Anywhere, we asked CEO Takahashi what is his strategy to continue to maintain its overwhelming domestic market share of 80% and he shared about RPAHD’s platform strategy and various business partnerships to accelerate the adoption of RPA as a movement and position RPAHD as the platform leader beyond just their own tools: “We offer not just our own RPA tools but also a ‘service platform’ called BizRobo! Station in collaboration with partners that provides other major products by Blue Prism, NICE, OpenSpan/Pega etc. For instance, NICE RPA is stable in call center operators where robots are used like shepherds. We are also certified in Feb 2018 as the first Blue Prism official training partner in Japan, which strengthens our earlier partnership when we signed on in July 2017 as Japan’s first Blue Prism Master Reseller.”
“We also provide education and training to support the creation of a system for business change and troubleshooting, and the nurturing and dispatching of RPA engineers to mitigate the issues related to the introduction and operation of RPA. Currently, we have over 100 partners and we would like to increase to over 400. We also have a sales network all over Japan from Hokkaido to Okinawa.”
“On 16 Nov 2018, we signed a business alliance with Japan Legaltech Association to develop solutions utilizing BizRobo to automate tasks such as digital migration of handwritten documents, automatic check of various contract documents, automation of the narrowing of case law search, etc.”
“In May 2018, SHIFT (TSE: 3697), which is engaged in software quality assurance and testing, announced that it began offering a RPA diagnostic and repair service ‘ROBOPIT’ that was jointly developed with our wholly-owned subsidiary RPA Technologies.”
“In Sep 2018, we collaborated with BrainPad (TSE: 3655) to jointly develop a business support package to support the automation of operations occurring in ecommerce management, including automation of sales forecast / price optimization model, automation of content check processing of PDF catalog contents, data aggregation necessary for measuring the effect of Web advertisement and customer acquisition, sale of products and follow-up after delivery, automating inventory tasks, competitor investigation by regularly crawling the competitors’ websites and check the items, price, delivery date, etc.”
“In Nov 2018, we jointly develop a RPA x AI Introduction support package plan with BrainPad that supports advanced utilization of AI added to BizRobo. Various AI plug-ins are provided to upgrade the scope of application, including (1) Image recognition plug-in to read image data and judge colors; (2) Language analysis plug-in which analyzes sentences such as data reviewed by humans and application documents, extracts and labels frequently-used words and summaries; (3) Character recognition plug-in to extracts character by calling AI OCR or external service; (4) Demand prediction plug-in which executes the demand forecast model by delivering parameters such as product data and delivery date.”
“Microsoft Japan has collaborated with RPAHD to jointly develop Robot-as-a-Service on Microsoft Azure and offer the solution since Aug 2018 not only to quickly and flexibility create and introduce a cloud environment necessary for RPA but also to develop new functions using AI.”
“On 10 Dec 2018, German ERP giant SAP announced that it will used RPAHD’s BizRobo tool to automate ERP input to SAP S/4 Hana Cloud and SAP ERP for the SME customer base. When introducing SAP ERP, not only does it takes a lot of manpower and time to input the initial data, but it is also required to maintain the freshness, precision and granularity of information in the ERP continuously after the operation is started. SAP commented that utilizing BizRobo will speed up the completion of introduction of its SAP solution from 6 months to about 9 weeks by reducing the work change and operation burden during implementation.”
“We have strengthened cooperation with AI companies with the aim for RPA to perform more complicated tasks. We have invested in ABEJA which offers AI solution for retailers and Cinnamon Inc which does AI for OCR (optical character reading). For instance, Cinnamon’s Flax Scanner is a RPA class 2 solution using the company’s own document reading engine Cinnamon AI which is widely used for handling documents such as contracts, resumes, sales progress reports, medical charts, handwritten application forms, real estate property information, receipts, etc. Rossa Voice, another flagship product from the company, is a voice dictation solution based on an automated word correction technology used for Flax Scanner. Nomura Holdings will use Rossa Voice to streamline the call center operations at their brokerage services. Aiming to improve their service quality and prevent possible defects, many brokerage firms record phone interactions between their agents and customers regarding deals. However, it would be impossible for humans to monitor in real-time or review all records. Rossa Voice can make these processes more efficient and labor-saving.”
CEO Takahashi summed up by emphasizing that RPA is not a tool that rob people of jobs but rather he envisioned humans working alongside the RPA digital workers while freeing up humans the time to do their most creative work: “According to a survey released by the US research firm Tractica in July 2017, it is predicted that RPA’s global market growth will accelerate from US$1.5bn in 2016 to US$51bn in 2025. The Japan RPA Association predicts that over 200 million white-collar work will be replaced by RPA worldwide by 2025. To emphasize, RPA is not a technology that robs people’s work. The biggest cost to the company is labor costs. There are a lot of inefficient operations left in indirect departments and service departments. To reduce the cost spent on huge amount of routine work and to have people take charge of more advanced and more creative work to increase productivity is the most effective cost management.”