NeoRMR

All operators of Artificial intelligence NeoRMR
will provide 24/7 customer support.

About NeoRMR

Interpret customers conversational contents and present the best answer candidate on the operator's screen. Operators can answer with just a click from among the possible answers.

  • Correspondence time reduction
  • Correspondence quality control
  • Training cost reduction
  • Customer satisfaction improvement
RMRdemo_operationside

NeoRMR Basic Functions

NeoRMR will learn QA combinations.

And when a new question is received, the system will choose the seemingly best answer candidate among those learned by NeoRMR. This is the basic function of NeoRMR

basic functions

Schedule For Full
Automated Operation

As AI learns, it will proceed towards full response automation.

schedule
  • Hybrid
    3 months
  • Escalation
    3 months
  • Full Automation
    6 months ~

Reason for Choosing NeoRMR: Cost &
Effectiveness

Labor cost savings of

approximately 97%
for data set during introduction

Labor cost reflecting learning data during operation

5〜10million JPY
in savings

Other companies NeoRMR
Introduction timeQA data set Large amount of data set creation/introduction is required Can be started from 0
During operationReflection of learning data Regular updating
Updating by data entry
Learn real-time during operation

Track Record of NeoRMR Introduction

About 100 large corporations have introduced NeoRMR into their systems.

About 100 large corporations introduced NeoRMR

NeoRMR Application Examples

Character AI Bot

Character AI Bot

AI chat system with character. Personalization possible.

Coach AI Bot

Coach AI Bot

Use AI as a partner to practice tasks that require repeated speech training such as operator skills etc.

Counselling AI Bot

Counselling AI Bot

Interview with AI as a psychological counselor. Record conversations to help diagnose patients.

Introduction Results in CS Services

(Scale of 150people)

Savings in labor hours

43,200hours

Converted into money

13,620,000JPY

Average Response Time(s) 20% Reduction
Number of cases handled
per hour
35% Increase
Number of monthly
acquisitions
20% Increase

The number of cases dealt by operators per hour increased.
The operator can reach the proficient level in a short time.

Introduction Steps

1 Step Hearing/Review 2 weeks
Start using in abou 2 months!
2 Step Application 1 day
3 Step Build environment/Issue account 2 weeks ~
4 Step Implementation of actual environment/Installation of initial data/Operation verification 4 weeks
5 Step Start Use/Management・Maintenance Start

Functions List

Support AI (NeoRMR)

<As of July 2019>
NeoRMR Core Functions
NeoRMR main system+NeoRMR internal API
NeoRMR None-core API
Ranker Learned Answer
Add evaluation value to candidate QA pair
Retrieval of answer candidates
Collections Collection Management
Registration of new Collection
Retrieval of Collection information list
Retrieval of specified Collection information
Updating of Collection information
Deletion of Collection information
Qas QA Management
CSV Batch Registraion
Deletion of all Qa information
Registration of new Qa
Retrieval of Qa information list
Retrieval of specified Qa information
Updating of Qa information
Deletion of Qa information
Updating of Qa tags
Searching Qa information using tags
Search exact match QA
Retrieval of Qa count
Dicts Dictionary Management
Batch registration of dictionary information
Registration of new dictionary information
Retrieval of dictionary information list
Search dictionary information
Retrieval of specified dictionary information
Updating of dictionary information
Deletion of dictionary information
Batch deletion of dictionary information
Dictionary test
2019 Release
NeoRMR multilingual support
NeoRMR core function in Indonesian
NeoRMR core function in English
NeoRMR core function in Vietnam
Master / Log / Authentication / AltGoPlatform / Tag
Master management API
Log measurement system / log reference API
OAuth2 authentication API
AltGoPlatformAPI
Function with QA tag
Tagging function for chat
Function to import data including QA_ID from another system
Function Extension
Autoreply when operator is absent
Scenario conversation with autoreply

2020 Release

Featured Functions

BOT Menu auto-generation function

Ability to automatically display in the BOT menu by ranking the answers that received many inquiries and high customer ratings

Other "voice input function" "FAQ added contents automatic generation" etc. will be released

Function Specifics

Regular Chatbots

Difference with Regular Chatbots

Regular Chatbots
  • It is necessary to construct a scenario in advance, and to register questions and answers exhaustively.
  • It cannot internally feedback (learn) new questions and answers such that an additional task to re-registering QA data occurs.
NeoRMR
  • A method of accumulating pairs of questions and answers in a question-and-answer format. In other words, you can start using it even with a small amount or without any QA pairs.
  • It cannot internally feedback (learn) new questions and answers such that an additional task to re-registering QA data occurs.
  • You can learn at high speed by using your own algorithm. (Quick updating of learning model)
  • New questions and answers can be learned within one second with a single click on the operator's screen, and learning can proceed almost in real time without awareness of the underlying complicated mechanism.
Regular Chatbots

Difference with Large-scale chat support system

Example:IBM Watson
  • Watson assistant: It is the same as regular chatbots where pre-registration is necessary.
  • Discovery: Create a learning model by registering and classifying the sentence data.
    It is a system which will return “which sentence is the closest?” when a query is sent to the model.
NeoRMR
  • A method of accumulating pairs of questions and answers in a question-and-answer format. In other words, you can start using it even with a small amount or without any QA pairs.
  • You can learn at high speed by using your own algorithm. (Quick updating of learning model)

Format of Use

API

API

For customers who are using chat systems, currently.

API + Interface

API + Interface

For customers who would like to introduce it into a new chat system.

Algorithm

Algorithm

For clients who would like to use it in an on-premise environment.

API DEMO

FAQ

A It can be introduced in any device. You can use it on PC, of course, and other devices such as smartphones, tablets or smart speakers etc.
NeoRMR is a chat operator AI support system that can be used effectively regardless of the industry. By using the dictionary function for specific industry terms, product names, and common names, it is possible to suggest candidates that are suitable as answers to customer questions.
The chat operator AI support system NeoRMR, can display answer candidates by learning past questions and their answers. Therefore, answer candidates are not displayed for questions with completely new contents. In that case, the operator will respond by entering and learning the answers.
The order of answer candidates is determined by an algorithm but can be changed manually.
The chat operator AI support system NeoRMR has an optimal plan according to the scale of use (number of users, number of inquiries, etc.). Please feel free to contact us.

Professors

Tomoyuki Nishita
Tomoyuki Nishita

The University of Tokyo
Professor Emeritus

Computer graphics, Visualization of natural objects and natural phenomena Lighting simulation, CAD system

Danushka Bollegal
Danushka Bollegal

University of Liverpool UK
Associate Professor

Natural language processing, Web data mining, Artificial intelligence, Statistical machine learning

Satoshi Kurihara
Satoshi Kurihara

Keio University
Professor

Artificial Intelligence, Complex Network Science, Computational Social Science

Ikushima Takahiro
Ikushima Takahiro

Mathematical Science Advanced Technology Laboratory Co., LTD
Professor

OS kernel technology Simulation system Data mining

Hashida Koiti
Hashida Koiti

Graduate School of Information Science and Technology
Professor

Natural language processing Service Informatics Distributed PDS

Ryohei Sasano
Ryohei Sasano

Nagoya University
Associate Professor

Natural language processing Computational linguistics

Satoshi Sekine
Satoshi Sekine

New York University
Associate Professor

Natural language processing, Information extraction, Linguistic knowledge acquisition, Named Entity Extraction

Phuong Le-Hong
Phuong Le-Hong

Hanoi University of Science and Technology
Professor

Natural language processing software authors such as Natural Language Processing Vitk, vnTokenizer, vnTagger

Junichi Yamagishi
Junichi Yamagishi

National Institute of Informatics
Professor

Speech information processing, Speech synthesis, Speaker recognition, Speaker verification, Machine learning, Biometrics

Matsubara Hitoshi
Matsubara Hitoshi

Future University Hakodate
Professor
Former Chairman of The Japanese Society for Artificial Intelligence

Artificial Intelligence, Gaming Informatics, Public Transport, Tourism Informatics

It is used by services and partners in various industries that frequently use communication.

RPA+AI

RPA+AI

It is possible to combine AI with RPA to maximize efficiency

Chat + AI

Chat + AI

Cost reduction & satisfaction improvement by combining AI with chat system

CS+AI

CS+AI

Adoption cost reduction and satisfaction improvement for call centers

Sales Agents

Sales Agents

Available for sale as a solution for your company

About

We are developing the Personal Artificial Intelligence, “PAI” that accelerates the realization of an autonomous society by eternally regenerating personal memories, reproducing intentions and maximizing individual values.

“NeoRMR” is one of reverse engineering roles of the brain and advanced conversation engine language field.

personal artificial intelligence

Possess an ACL / COLING Adopted Technology

World AI language processing society adopted technology
Ranked 7th in the world Japan Venture AI Research Company Rating

Possess an ACL / COLING Adopted Technology

Media

  • 日本経済新聞
  • TC
  • 毎日新聞社
  • 日テレ
  • c net
  • SankeiBiz
  • TV Tokyo
  • 東洋経済
  • 朝日新聞社
  • WIRED
  • THE BRIDGE
  • Yahoo!
  • tv asahi
  • President Online
  • 日経XTECH
  • ITmedia
  • Diamond Online
  • The Japan Times
  • NHK Eテレ

Company

Location

HEAD OFFICE

703, SENQ Roppongi, Shin-Roppongi bldg.
7-15-7 Roppongi, Minato-ku, Tokyo, Japan 106-0032

 

OVERSEAS

Vietnam Tech Labo:
11F, Lucky Building, 81 Tran thai Tong street, Hanoi, Vietnam

GROUP COMPANY

・alt Inc.