Executive Summary

The company offers emergency roadside assistance in North America. They partnered with ITTStar to develop a lucrative solution to automate IT processes & business operations to deliver customer service. Leveraging the industry leading machine learning capabilities of AWS, ITTStar enabled the customer to deliver valuable business impact by adding intelligence & personalization along with predictive analysis to enhance the overall customer experience.

Customer challenge

The customer was aiming to automate workflows to deliver enhanced roadside assistance to its customers. Their objective was to improve the overall operational efficiency of their call center in order to provide quick actionable assistance to their customers who are stuck on the road.

To attain the objective, the customer wanted to develop a unique model using technology to understand the caller-agent dynamics, taking cues on keywords, topics,entities, and sentiment in real-time during the call between the agent and the customer.

Partner Solution

To solve this business challenge, the customer engaged with ITTStar who built an AI-driven model to help the customer derive caller insights to provide customized approach to the agents thus, ensuring enhanced customer delivery and satisfaction. ITTStar approached this innovative project in four distinct phases:

• Phase 1: Discovery – Identified all sources and types of data available,including the quality, relevance, and usability of audio files between caller & agent.

• Phase 2: Analysis & Recommendation – ITTStar recommended an AI-driven model to analyze the audio files to produce insights on keywords, topics, entities, and sentiments from the text transcripts.

• Phase 3: Defining methodology – ITTStar laid out the methodology and services that would be used to decipher call center audio files, run sentiment analysis, and visualize analytics.

• Phase 4: Defining methodology – ITTStar built this unique solution using various AWS services to achieve the desired outcomes: Stored the audio files in Amazon S3 bucket, triggered an AWS Lambda function to invoke AWS Step Functions to point the Amazon Transcribe service to the bucket destination to create transcription jobs
  Enabled AWS Step functions to call Amazon Comprehend to analyze transcription text to produce sentiment analysis and keywords and stored in Athena
  Visualized sentiment analysis and key phrases in Amazon QuickSight based on the data stored in Athena table.

Why AWS?

Innovations in new technologies such as AI & Machine Learning are key offerings from AWS enabling greater data insights, faster decision making & greater agility.The effective cost models available, are very lucrative for small businesses to transform their traditional practices, re-imagine customer experience and rethink the future and power of technology to enable customer success. Amazon Comprehend & Amazon Transcribe were used to extract key elements and insights of customer calls. API Gateway was used to integrate with the customer’s existing CRM and Amazon QuickSight for complete visualization of all data points.

Impact & Results

ITTStar’s unique solution transformed the customer’s business functioning in the following ways: :

• Enhanced efficiency and overall performance of call centre & service quality to its customers
• Delivered quality insights using sentiment analysis for actionable solutions and customer interactions real-time
• 15% increase in customer satisfaction score & enhanced customer engagement
• Improved agent’s day to day experience, lowering churn & resulting in lower training and hiring costs
• Delivered agent performance statistics which provided insights to managers and other leadership for performance assessment of employees