PKSHA Technology List of Products and Services

There is no need to aim for perfect automated responses from the initial implementation. We will start with human support first. The data accumulated from human support will be transformed into knowledge by AI (STEP 2), and using that knowledge, we will begin automated responses (STEP 3), creating a "virtuous cycle." This journey design allows us to expand the automation area without difficulty. *The complete details of the fail-safe implementation steps are explained in the final chapter of the document.

• Chatbots

We introduced AI to predict the large volume of inquiries during the implementation of the new system and to improve the complicated reception system. We trained the AI help desk using approximately 5,600 accumulated inquiry logs to create a system that presents highly accurate FAQs. We are continuously expanding the area of automation by utilizing analysis from generative AI. *The secrets to successfully leveraging data for large-scale implementation are detailed in the materials.

• Chatbots

Employees expressed dissatisfaction with the difficulty of reaching the help desk by phone, and the neglect of outdated FAQs became an issue. We implemented an AI help desk and established a clear process of "first the bot, if unresolved then human chat." As a result, we successfully closed the phone support completely and consolidated 100% of inquiries into Teams. *For specific design details that achieved zero phone support, please refer to the case study in the materials.

• Chatbots

In sales environments that handle a wide range of products, the challenge was that it was difficult to memorize the vast amount of manuals, leading to time-consuming information searches. By integrating Microsoft Teams with an AI help desk, we established an easily accessible inquiry window. AI provides immediate responses, while only a human team addresses complex questions, achieving a sales digital transformation that does not keep customers waiting. *Details of this successful case that enhanced productivity in the sales field are available in the materials.

• Chatbots

The main reason for the ossification of the internal FAQ site is the lack of human resources dedicated to updates. The AI help desk learns from manuals and daily response logs, automatically creating and updating FAQs. As inquiry logs accumulate, the quality of automated responses improves, allowing you to maintain the most up-to-date information without the hassle of management. *For information on how to achieve maintenance-free operation, please refer to the steps for utilizing the materials.

• Chatbots

Inquiry management in Excel faces many issues, such as file regression, data bloat, and insufficient notifications. The AI help desk shares information in real-time online and centrally manages "who," "when," and "how" responses were made. The notification feature among team members prevents missed responses and establishes an efficient support system for the entire team. *Specific features that alleviate the stress of management tasks are explained in detail in the materials.

• Chatbots

Before the implementation, we were struggling to find manuals and wasting time not knowing the contact points, but after the implementation, everything changes. By consolidating all initial inquiries into the AI help desk, simple questions can be resolved instantly through chat. A refined flow is established where only complex issues that cannot be resolved by AI are smoothly handed over to the responsible personnel. *The before-and-after diagram illustrating how the business changes is the main content of the materials.

• Chatbots

The main difference from AI chatbots is that it goes beyond mere text responses to offer comprehensive features. AI help desks encompass automated responses, integration with human responses, ticket management, detailed analytics, and automatic FAQ generation. They integrate various interactions related to internal inquiries, achieving both efficiency and quality improvement simultaneously. *The differences between the two tools and points for their appropriate use are clearly explained in the comparison table in the materials.*

• Chatbots

In recent years, the trend of generative AI has shifted from standalone use to the integration of internal documents. AI help desks utilizing RAG (Retrieval-Augmented Generation) technology serve as gateways to extract necessary information from vast knowledge data. By accumulating and managing the insights of internal experts, and allowing AI to leverage this information, advanced knowledge management is achieved. *Please check the materials for an advanced knowledge utilization experience using generative AI.*

• Chatbots

There are many cases where people cannot find the necessary information even after searching for it, or they waste time not knowing where to make inquiries. The AI Help Desk instantly provides appropriate information sources and contact points based on the content of the questions. This eliminates "information lost" within the company to zero and thoroughly eliminates wasted time on inquiries. *The experience flow that alleviates the stress of information retrieval is detailed in the materials.

• Chatbots

You will be able to accumulate and utilize internal documents and chat exchanges as knowledge. By using the AI help desk, it is possible to automatically generate FAQs from dialogue logs. The system can capture the insights of experts and easily build a flow to redistribute them as assets for the entire organization. *For a specific list of features to capitalize on knowledge, please refer to the materials.

• Chatbots

There is no longer a need for the person in charge to stop and answer frequently asked questions from new employees who lack knowledge. The AI help desk has learned from past inquiries and will automatically provide appropriate answers to rules and procedures that new employees may find confusing. This saves time for both the instructors and the learners, facilitating a smooth onboarding process. *Please refer to the materials for information on the form of knowledge transfer that contributes to the development of young talent.*

• Chatbots

The AI help desk is a system that automatically responds to inquiries both inside and outside the company using AI. It generates answers based on pre-set FAQs, internal documents, and past conversation logs. As data accumulates, the accuracy of the responses improves, providing a help desk environment that "grows" the more it is used. *Details of the AI learning mechanism and the response process are illustrated in the accompanying materials.*

• Chatbots

In modern times, communication methods have diversified, and inquiries from multiple channels are complicating business operations. By implementing an AI help desk, you can consolidate the disparate inquiry channels and accumulate all data in a usable format. Let's evolve from the daily grind of being overwhelmed by phone calls and emails without a structured flow to a data-driven, smart operation. *We provide solutions for organizing complicated business operations in our materials.*

• Chatbots

As the number of inquiries increases, both the time and psychological burden accumulate, leading to an increase in communication costs. The AI help desk learns from past interactions and documents, continuously providing automatic responses to similar questions. By eliminating the waste of "rechecking" that occurs when knowledge is not shared within the organization, it maximizes the use of the organization's insights. *The specific mechanisms for cost reduction are explained in the analytical data within the materials.

• Chatbots

If the clarification of unclear points within the company does not go smoothly, it will affect not only that employee but also customers and the entire team. Implementing an AI help desk to optimize the environment is not just about efficiency; it is an investment to increase the speed of business execution. By bringing the waiting time for responses close to zero, it becomes possible to elevate the productivity of the entire organization. *Please refer to the materials for the deeper reasons why optimizing the inquiry environment is necessary now.

• Chatbots

At the end and beginning of the month, as well as during the financial closing period, inquiries related to expense reimbursements and payments tend to concentrate in the accounting department, which can disrupt their primary tasks. With an AI help desk, common questions such as reimbursement rules can be answered automatically 24/7. This eliminates the overlap between inquiry responses and payment tasks, creating an environment where the accounting team can focus on their error-sensitive work. *If you are considering measures for handling inquiries during peak periods, please refer to the case studies in the provided materials.

• Chatbots

The Human Resources and General Affairs Department receives a wide range of inquiries related to attendance, regulations, equipment, and employee benefits on a regular basis. In cases where the point of contact is unclear, it is common for people to say, "Let's consult General Affairs for now," which tends to complicate the response process. By establishing an AI help desk as the point of contact, the AI can instantly classify and respond to a broad range of questions, significantly reducing the psychological burden on the staff. *For details on methods to dramatically reduce the burden on management departments, please refer to the materials.

• Chatbots

The information systems department, responsible for the implementation and management of the system, is constantly receiving inquiries about operation methods and malfunctions. By introducing an AI help desk, AI will take over the frequently occurring IT-related help desk tasks. As a result, the staff can focus on high-value tasks that they should be concentrating on, such as system maintenance and operation. *We have published specific tips for improving the efficiency of the information systems department in a document.

• Chatbots

The vast number of inquiries received from within the company is essential for business operations, but it can also tend to pressure the core tasks. As a way to eliminate the waste in communication between those making inquiries and those receiving them, AI help desks are currently gaining attention. This document provides a detailed explanation of how business operations can change by utilizing advanced generative AI technology, covering everything from basic knowledge to specific examples. *If you want to build an ideal work environment, please download this document.*

• Chatbots

We will explain the successful steps to grow from a small start in individual departments to a company-wide common inquiry contact point alongside phone and email. We will begin with management departments such as IT and HR, where automation opportunities are significant, and expand while increasing internal awareness. Over a period of 12 to 24 months, we will develop into a corporate integrated contact point, ultimately allowing for expansion into front departments such as sales planning. Through supportive assistance tailored to this phase, we will achieve the minimization of inquiry response costs on a company-wide scale. *We will provide a detailed roadmap for successful company-wide deployment in the materials.

• Chatbots

We propose a method to eliminate the hassle of "creating FAQs," which has been the biggest barrier to the introduction of conventional chatbots, allowing for a speedy start to operations. By simply setting the documents you want to search and installing the Teams app, you can begin operations as soon as the same day. First, by operating a human chat while allowing AI to automatically generate FAQs, you can gradually expand the scope of automated resolutions. This "smart operation while running" approach dramatically shortens the lead time for implementation. *An example schedule for starting operations in as little as one month is currently available in the materials.

• Chatbots

We will introduce how PKSHA, a specialized package, addresses the risks of incorporating generative AI within a company. While tools like Copilot and in-house RAG offer high flexibility, they often present challenges in controlling hallucinations (incorrect answers) and managing operations across various departments. PKSHA packages FAQ, document search, and human support, minimizing the risk of incorrect answers while being usable across all departments. Its intuitive UI allows for operation beyond just the IT department, making it well-suited for company-wide digital transformation initiatives. *Detailed feature comparison tables with similar services are included in the materials.*

• Chatbots

We will explain the decisive differences between typical bots that tend to be unused even after implementation and PKSHA's solution. Typical chatbots are abandoned if they cannot resolve issues on their own, but PKSHA ensures resolution through strong collaboration with human chat support. Additionally, there is no need for prior preparation of FAQs, as there is a mechanism to automatically generate knowledge from human support logs, allowing users to experience effectiveness from the very first day of implementation. The fact that tools are not scattered and that operations can be completed on Teams is also a crucial factor for continued use by employees. *We provide a detailed comparison of the criteria for choosing a chatbot that won't fail in our materials.

• Chatbots

We will unravel the specific financial benefits brought about by the automation of inquiry responses based on actual calculation examples. In a company that receives 50,000 inquiries per month, if 50% of them can be automated, it would result in a reduction of 8,400 hours of labor per month. When converted to the hourly wage of full-time employees, this translates to an estimated cost benefit of approximately 350 million yen annually. Shifting the reduced resources to more creative and high-value-added tasks will contribute to the growth of the organization. *We have published indicators for estimating the reduction effects for your company in the materials.

• Chatbots

By analyzing past inquiry data and distributing information before company-wide issues arise, we can prevent the occurrence of inquiries themselves. We have implemented a "Proactive Agent" feature that proactively delivers notifications from AI agents to employees based on log analysis. For example, when there are signs of a system failure, we can prevent a flood of similar inquiries by distributing solutions in advance. Instead of a passive approach, we enhance the productivity of the management department through a proactive help desk operation. *For effective ways to utilize the "notification distribution" as a remedy for reducing inquiries, please refer to the materials.

• Chatbots

By converting inquiries into tickets and visualizing the person in charge and progress status in a list, we optimize help desk operations. Inquiries can be ticketed on Teams, allowing for management of status, deadlines, and responsible persons in a list format, which helps prevent oversight and bias in responses. Administrators can check overall KPIs on a dedicated dashboard, making it possible to streamline the management of response tasks. Frontline staff can complete their work on Teams, which they are already familiar with, resulting in a very low psychological burden for implementation. *Check the list of features that maximize help desk management efficiency in the materials.

• Chatbots

We introduce a groundbreaking feature where polite responses from human chat are directly accumulated as the next automatic response knowledge. From the conversation history in human chat, the generative AI automatically creates pairs of "questions" and "answers," proposing them to the administrator as potential FAQs. The administrator can check the content with a single click and register it for the AI agent to learn. By repeating this cycle, the more it is operated, the smarter the AI becomes, naturally reducing the number of cases that require human intervention. *For details on the mechanism of "knowledge generation" where AI self-develops, please refer to the materials.

• Chatbots

We will build an environment where AI can always reference new information by integrating with SharePoint Online used within the company. It is possible to synchronize the AI help desk with a specified folder, allowing for immediate reflection of any file additions or updates. There is no need to organize all files, and the ability to start small from a specific folder for public use is a significant advantage. This will prevent the double management of manuals and enhance the accuracy of automated responses while maintaining the freshness of information. *Specific settings and operational images for SharePoint integration are illustrated in the materials.

• Chatbots

We introduce the "Document Agent," which allows AI to read existing internal manuals and generate appropriate answers to questions. By simply registering documents such as PDFs and PowerPoints, the AI agent automatically analyzes the content and generates responses. This eliminates the hassle of creating FAQs one by one, allowing for a quick expansion of automated resolution tasks while keeping management workload low. Since the name of the source document is also provided, employees can confidently utilize the information. *Tips for improving work efficiency using document search are currently available in our materials.

• Chatbots

We will explain an advanced response engine that combines our unique AI, which utilizes big data from 55 million users, with cutting-edge generative AI. PKSHA's engine is equipped with a dictionary based on over 55 million user data, allowing it to understand variations in language and industry-specific phrases. For example, it can accurately derive responses to diverse expressions such as "I forgot my password" and "I don't know my PW." This in-house developed AI agent guarantees a top-class response accuracy of 90% to 95%. *We will provide detailed information on the mechanism that achieves this remarkable response accuracy in our materials.

• Chatbots

By not relying solely on AI and smoothly collaborating with humans, we lead various inquiries to a 100% resolution. PKSHA is based on the premise of Human-in-the-Loop, combining "FAQ search," "document search," and "human support." Frequently asked questions are answered immediately by AI, while complex individual cases are seamlessly handed over to human chat, ensuring employees are not kept waiting. This three-step structure maximizes answer accuracy and coverage, simultaneously improving employee satisfaction and productivity. *Details of the "three steps" that achieve an astonishing resolution rate are explained in the materials.*

• Chatbots

Many leading enterprise companies in Japan, regardless of industry, are adopting PKSHA's solutions. Among the companies that have implemented these solutions are major telecommunications firms such as NTT East and Docomo, as well as global companies like Toyota, Mitsubishi Corporation, and Marubeni. Additionally, they are widely utilized in sectors such as finance, construction, real estate, and educational institutions, where strict security and accuracy are required. The experience of providing algorithms to over 4,400 companies has built trust in meeting advanced requirements. *Case studies of companies in your industry are also available. Please see the detailed materials.

• Chatbots

We will introduce specific quantitative effects and best practices for operations utilizing advanced AI at a certain bank. Out of a total of 901 inquiries, the FAQ agent quickly resolved approximately 41% on its own. For cases that could not be resolved, the document agent generated answers from the manual, covering an additional approximately 30%. Ultimately, only a small portion of inquiries were connected to a human chat, achieving significant operational efficiency as a successful case. *Download the case study document now for detailed figures and operational flow.*

• Chatbots

We propose a method to complete all inquiry tasks on Microsoft Teams, which employees use on a daily basis. Since this service is provided as a Teams app, employees do not need to learn new operations, resulting in a very low psychological barrier. Through chat, the AI agent explores information, and if it cannot resolve the issue, it seamlessly connects to human support. Management can also manage tickets and perform analysis on Teams, significantly reducing the effort required for tool adoption. *Details of the business transformation through Teams integration are illustrated in the materials.

• Chatbots

We will solve the issue of "the FAQ not being updated despite the effort put into its implementation" with an AI-powered automatic maintenance feature. The traditional system required manual updates for articles, leading to a backlog of unresolved issues and a lack of improvement. With PKSHA's system, the generative AI automatically creates and suggests FAQ candidates from dialogue logs, significantly reducing the maintenance workload for administrators. Designed with a focus on employee experience, it ensures that new knowledge is always provided, preventing obsolescence. *For more details on the automation system that makes operations easier, please refer to the documentation.

• Chatbots

We fundamentally solve the issue of "dependency on specific individuals," where information is scattered across multiple tools. In many companies, information is dispersed across SharePoint, emails, manuals, and other sources, making it difficult to search. PKSHA's AI Help Desk consolidates various inquiry channels into a chat, allowing AI to instantly search for information. This eliminates the time employees spend getting lost at the "service desk," significantly improving productivity across the organization. *Would you like to find hints to solve your company's help desk challenges in our materials?*

• Chatbots

PKSHA's AI technology, adopted by many major companies, boasts overwhelming market share and reliability. PKSHA Technology is listed on the Prime Market of the Tokyo Stock Exchange and is a specialist with a track record of over 2,600 implementations. In the market research for the fiscal year 2023, it achieved a high market share of 18.9% in the chatbot market and 28.6% in the voicebot market. Our experience in providing algorithms to over 4,400 companies strongly supports your company's digital transformation (DX). *Details of the technology that supports our overwhelming implementation track record are explained in the materials.

• Chatbots

We introduce a new platform to maximize the vast knowledge buried within companies using AI. This solution aims to solve problems without compromising user experience by appropriately separating issues that should be handled by AI and those that require human intervention. A key feature is that the AI agent automatically formalizes tacit knowledge from logs of human interactions, dramatically reducing the burden on administrators. Going beyond traditional FAQ management, we provide an interactive system where humans and AI co-evolve, accelerating corporate growth. *Please check the materials for a comprehensive overview of future knowledge management.*

• Chatbots

"PKSHA AI Helpdesk" realizes the "improvement of the inquiry experience" and "automation of responses" through a business process that combines AI and humans. The AI agent provides immediate answers through FAQs and generates responses by reading documents. It can also call upon the appropriate person, learn from the resolved answers, and apply that knowledge autonomously. By achieving a seamless collaboration between AI and humans, it builds a system that is truly used on-site, contributing to increased productivity and improved employee experience (EX).

• Chatbots