← Back to works
03 / Selected work

MentorAI Call Transcription & Analysis

An AI platform for the automatic transcription, analysis and quality assessment of sales managers' calls

MentorAI — project cover
Company NDA
Role Senior Product Designer
Period 2026

Project Overview

Development of an intelligent platform for call analysis featuring automatic transcription, sales performance evaluation, and manager training.

The system uses NLP and machine learning to identify strengths and areas for improvement in managers' conversations.

Problem

The company's sales department handled over 500 calls a day, but listening to and evaluating them took up 80% of managers' time. Only 5% of calls were analysed manually, feedback was delayed by 3–5 days, and the objectivity of the evaluation depended on the manager's subjective opinion.

Solution

Designed an AI platform with real‑time automatic transcription and analysis across six modules: establishing contact, identifying needs, product presentation, handling objections, closing the deal, and professionalism.

Implemented a colour‑coded quality rating system (from 0/10 to 10/10), detailed visualisation of strengths and areas for improvement, and integration with CRM.

Visuals

MentorAI — main dashboard
01 Main dashboard — overview of recent calls and team performance.
Calls list view
02 Calls — filterable list with quality scores and status.
Call detail — full view with transcript and analysis
03 Call detail — full transcript, six-module evaluation and AI insights.

Results

  1. 01

    100% of calls are analysed automatically, instead of 5%

  2. 02

    The time required to analyse a single call has been reduced from 15 minutes to 30 seconds

  3. 03

    The conversion rate of calls into deals increased from 12% to 23%

  4. 04

    The average check increased by 18% thanks to improved sales techniques

  5. 05

    A monthly time savings of 120 hours for managers

  6. 06

    The training period for new managers has been reduced from 3 months to 3 weeks