FRONTEO to provide compliance check PoC service for financial institutions with Advanced Media

2017-06-29 14:02


Verification process for identifying inappropriate responses extracted from actual phone call recordings

Demonstrations available at the FRONTEO booth during AI EXPO from June 28-30, 2017

TOKYO, June 27, 2017 (GLOBE NEWSWIRE) --  FRONTEO, Inc. (“FRONTEO”) (NASDAQ:FTEO) (TSE:2158), a leading provider of artificial intelligence (“AI”) based big data analysis services, and Advanced Media, Inc. (“Advanced Media”) (TSE:3773), a leading provider of product solutions utilizing voice recognition and voice recognition UI (user interface)/UX (user experience), will begin a Proof of Concept (“PoC”) trial run conducting compliance checks for financial trades placed over the phone.

This service aims to provide an accurate verification system by constructing a process utilizing actual phone call recordings and seeks to allow financial institutions conducting sales activities by phone to easily detect inappropriate responses that may breach laws such as the Financial Instruments and Exchange Act.

The compliance check service for phone call recordings service will seek to do as follows:

1.      Conduct tuning services to add customer-specific vocabulary data, build language models, and construct acoustic models for Advanced Media’s voice recognition engine AmiVoice.

2.      Build a voice recognition system with an embedded and tuned voice recognition engine within the financial institution’s systems.  Training on the various settings and methods for use will also be offered. The system will input recorded phone conversations and convert conversations to text within the actual environment.

3.      Input content that should be detected through analysis as training data for FRONTEO’s artificial intelligence (“AI”), KIBIT, to teach it to identify situations including:

  • Communication between a sales representative and customer that might infringe on compliance standards

  • Transactions that have the possibility of deviating from company or business policies and rules

4.      Analyze recorded phone conversations that have been converted to text with KIBIT. Verify KIBIT’s accuracy in detecting and scoring inappropriate responses. Adjust tuning as needed.

Both FRONTEO and Advanced Media have an extensive track record of working with financial institutions and believe they possess the compliance knowledge necessary to provide a one-stop process to achieve financial institutions’ compliance goals. Furthermore, FRONTEO and Advanced Media’s services include converting conversation records into text, tuning to improve AI analysis accuracy, and training on application and engine usage, allowing a smooth transition to the service beginning with early implementation or from verification. The verification period is estimated to be approximately three to four months for a standard workload.

The AI engine, KIBIT, has showcased an ability to correctly identify and score compliance-related verifications even when text was incorrect by understanding the surrounding context. Thus, a combination of voice recognition and text analysis technology makes it possible to find context for analysis even if the sentence text is not correct.

Currently, every financial institution is engaged in fulfilling its fiduciary duties. Both FRONTEO and Advanced Media believe that the implementation of their compliance check service will greatly improves the quality and efficiency of identifying high-risk records . Furthermore, as employees improve their behavior, customer care can also improve. FRONTEO and Advanced Media will continue to offer this service while exploring expansions in its application, such as using it for responding to customer inquiries.

The compliance check PoC trial run for phone call recordings will be exhibited at the AI Expo scheduled to be held from June 28 through June 30 at Tokyo Big Sight. For more information on AI Expo, please visit

Name: 1st AI EXPO: Artificial Intelligence Exhibition & Conference
Dates: Wednesday, June 28 - Friday, June 30, 2017
Venue: Tokyo Big Sight
Booth: 32-39 (East Hall 6)