RightTouch Launches 'QANT Coach (β)', an AI-Powered Automatic Call Quality Evaluation Solution

RightTouch Inc. (Headquarters: Shinagawa-ku, Tokyo; Representatives: Shuhei Nomura / Daito Nagasaki; hereinafter "RightTouch"), a provider of AI-powered contact center platforms for enterprises, has launched a new product called 'QANT Coach (β)'. This solution leverages generative AI to automatically and uniformly evaluate the quality of all customer interactions in contact centers.

The product addresses the longstanding challenge where only a fraction of customer interactions could be evaluated due to supervisor (SV) workload limitations. Now, generative AI evaluates all interaction data—including email, phone, and chat—using a unified standard. This enables operations to rapidly cycle through feedback and learning processes based on evaluation results.

In the future, RightTouch aims to apply this same platform not only to human operators but also to AI operators, ultimately realizing a 'self-evolving contact center' where both humans and AI continuously improve under the same quality standards.

Development Background: Three Barriers in Interaction Quality Evaluation

Maintaining and improving interaction quality is a critical issue in contact centers, directly impacting customer experience (CX). However, quality evaluation processes face three structural challenges:

Challenge 1: Reliance on manual evaluation and feedback based on limited samples

Interaction quality evaluation has traditionally relied on supervisors (SVs) manually reviewing and scoring individual interactions. Due to the high labor cost per evaluation and limited SV resources, only a small fraction of total interactions can be assessed. As a result, feedback is routinely based on sampled interactions, leading to biased evaluations, missed critical issues, insufficient feedback, and low operator satisfaction.

Challenge 2: Inconsistent evaluation standards across evaluators

Interaction quality involves qualitative judgment, meaning the same interaction may receive different scores or feedback depending on the evaluator. Without consistent calibration across SVs, organizations struggle to maintain uniform evaluation standards, undermining the consistency of results. Operators are thus placed in a situation where outcomes depend on 'who evaluates them,' further reducing their trust in feedback.

Challenge 3: Overburdened with evaluation, leaving no time for essential training

The evaluation process itself consumes significant resources, inevitably reducing the frequency of feedback. As a result, the most critical aspect—training and learning cycles—often gets neglected. Without a mechanism linking evaluation results to operator improvement, organizations find it difficult to sustainably enhance interaction quality.

At the root of these three challenges is an operational design predicated on 'human-led evaluation.' A fundamental solution requires scaling the evaluation process and accelerating the cycle from evaluation to learning.

What is QANT Coach (β)?

'QANT Coach' is a quality evaluation product that uses generative AI to uniformly assess 100% of interaction data against standardized criteria, enabling rapid learning cycles starting from evaluation results. The first version is being launched as an optional feature of 'QANT VoC'.

Previously

Now (with QANT Coach)

Evaluation Scope

Only a subset of interactions (sampling)

100% of interactions automatically evaluated

Evaluation Criteria

Varies by evaluator

Uniform and consistent AI-driven evaluation

Learning Cycle

Overwhelmed by evaluation; low feedback frequency

Accelerated learning with weekly/daily feedback

Supported Channels

Primarily phone

Cross-channel: email, phone, chat

＜Value Propositions＞

① Full-volume automatic evaluation of all interactions

Simply input data and generative AI automatically performs evaluations. Free from SV workload constraints, all interaction data—including AI-handled emails, calls, and chats—can now be evaluated.

② Uniform evaluation based on standardized criteria

Optimal LLMs are applied per evaluation item, ensuring consistent and fair assessments under unified standards. This eliminates evaluator bias and delivers feedback that operators find credible and actionable.

③ Accelerated learning cycles

Weekly or even daily feedback becomes feasible, dramatically speeding up the evaluation → improvement → learning cycle. Supervisors can focus on targeted reviews and comments, while operators take ownership of their own development.

④ Flexible evaluation items and report design

Users can immediately start operations using preset evaluation items such as 'need identification,' 'clarity of explanation,' 'problem-solving approach,' and 'guidance that builds trust.' Additionally, the system supports custom evaluation items and full customization of component layout and output frequency on evaluation sheets.

Future Roadmap

'QANT Coach' will continue enhancing its functionality ahead of a planned full release in 2026, incorporating real-world operational insights from customers. The product will evolve to offer greater flexibility in evaluation items and reporting, supporting organization-wide deployment and enabling operators to conduct self-reflection. RightTouch also plans to introduce features that directly support operator learning experiences based on evaluation outcomes.

Looking ahead, the company intends to extend the application of this platform beyond human operators to AI operators, aiming to realize contact centers where both humans and AI evolve continuously under the same quality standards.

Executive Comment

Daito Nagasaki, Representative Director, RightTouch Inc.

"Interaction quality is the core of contact center operations, yet until now, most organizations have been stuck in a cycle where 'SVs evaluate a few interactions and stop.' There are inherent limits to both evaluation volume and consistency, leaving little bandwidth for what should be the most important function—training. This product directly confronts that challenge with generative AI.

Originally offered as a feature within 'QANT VoC,' we are now launching it as a standalone product in response to strong customer demand. With the ability to uniformly evaluate 100% of interactions, we can finally establish a fast feedback-to-learning cycle. Moreover, this very mechanism can be applied to the self-evolution of AI operators. We aim to build a service environment where both humans and AI grow continuously under shared quality standards."

Webinar: VoC × Interaction Quality Evaluation (June 24–26)

June 24 (Wed) – June 26 (Fri): "How AI is Transforming Call Center Quality Evaluation and Operator Development" (Sessions at 14:00–14:50 JST)

▼ Register here:

https://attendee.bizibl.tv/sessions/seSLme44BlRo

About QANT

'QANT' is a product suite that comprehensively supports AI implementation across customer support operations and touchpoints, enabling the realization of AI-powered contact centers. It connects and optimizes end-to-end workflows. Labor-intensive tasks such as issue analysis, planning, and knowledge creation are automated with AI, while human teams focus on content verification, decision-making, and interpersonal communication.