BrandPost: A Customer Intelligence Platform on AWS for a Superior Customer Experience

BrandPost: A Customer Intelligence Platform on AWS for a Superior Customer Experience

There’s a solution to accelerate customer insights by better analyzing and even anticipating client needs.


Customer experience has become a strategic priority on the CEO’s agenda. Enterprises now consider customer delight to be a differentiator in the marketplace.

Sentient technology is critical to serve customers better and be responsive to their needs. Enterprises must continuously learn, evolve, and develop a mechanism of interaction among agents, employees, and other constituents in this journey. At the same time, systems need to sense and respond to patterns and personas for enhanced interactions leading to a superior customer experience.

Digital consumers seek personalized service tailored to their unique needs. Accordingly, the customer experience needs to be sentient – where processes adapt to shifts in behavior, and employees are empowered with customer insights to make timely and informed decisions.

The Infosys Customer Intelligence Platform (CIP), built on Amazon Web Services (AWS), is a solution to accelerate data ingestion, data processing, data modeling, and data analytics. This smart platform:

  • leverages data such as events, interactions, external, telemetric data, IoT
  • builds knowledge graphs around customer identity, products, financial solution needs, risk, preferences
  • applies artificial intelligence (AI) or machine learning (ML) to understand data from multiple dimensions

It offers a unified experience providing users with hyper-contextual insights, omnichannel experiences, hyper-personalization and a 360-degree view of the customer and product.

The Infosys Customer Intelligence Platform can anticipate the needs of applicants and businesses including real-time product recommendations. In addition, the platform can understand an individual’s needs based on existing data to sense, respond, and automate routine decisions. These include moving the application to the next stage, requesting additional data, wearable data-based decisions, underwriting (UW) decisions using ML, experience modification rating (EMR), utilizing UW requirements grid, calculating premium, calculating risk using past data.

The platform can also reimagine and bring zero latency in key business processes, such as completing the quote process in a few steps, persona-based application handling to reduce the completion duration of complex cases.

The platform has three building blocks:


  1. Sense signals in real-time: Capability to gather physical data of the event in real-time
  2. Unified semantics: Foundational capability to sense data and access from core systems through unified semantics and ontology
  3. Co-relate and mind map: Hyper-contextual physical and digital data about an event, such as an accident


  1. Digital brain: Data + algorithms + learning to drive intuitive decisions— such as AI-based claims for the adjuster persona
  2. Knowledge graph: A graph database and compute, mapping the many-to-many relationships between the entity and other components such as customer, engagement, interactions, product, policyholder, claims adjuster
  3. AI services: ML capability to analyze data, images, documents, and unstructured information through microbots to help influence decisions

Experience orchestration

  1. Sentient experience configurator: Capability to anticipate and extract the most important feature based on current customer needs
  2. Personalized campaigns: Across channels including mobile, web, email subject, email, social media, offers, conversational commerce, and video
  3. Automate the business process: Adaptive supply chain, promotion, and channel monetization

A sample use case

An auto insurance policy holder collides with another vehicle while driving his car. He notifies the insurance company via the assistant on his smartphone. He shares accident details via voice or keyboard, and uploads images. The claim is analyzed, and the estimate is generated.

The claim is validated for fraud, and the claims adjuster is notified about the system-generated estimates. After review, the claims adjuster shares the estimates with the policyholder. Additionally, he is notified about service centers in the neighborhood.

Technical workflow leveraging CIP

  • Mobile app user authenticates credentials using AWS Cognito and gets a token on successful authentication
  • Token is passed along with API call and authorizes in API gateway. When API is authorized, it triggers the respective Lambda function (via API gateway)
  • Lambda function communicates either with EC2 instance, which hosts Graph DB server or Amazon S3 bucket and executes business logic in Lambda function

CIP leverages machine learning to identify fraudulent claims

The platform uses ML models for processes such as auto damage detection, auto policy recommendations, and fraud detection models, which are trained with large datasets. Platform users may also leverage bring-your-own-model (BYOM) approaches to implement any business use case, which can be supported by the Analyze module.

CIP drives meaningful information through knowledge graph

The platform captures key events and builds relationships to derive meaningful information, and enables intelligent decision making from:

  • Customer history and social data
  • Customer vehicle information
  • Geo-events data
  • First-party and third-party data

Infosys Customer Intelligence Platform is a modular offering to enable a next-gen customer experience for enterprises. Infosys and AWS bring capabilities together to drive customers’ business and IT priorities through comprehensive cloud offerings delivered through Infosys Cobalt.

About the Authors

Arav Narasimhamurthy, is the Senior Principal Technology Architect – Data and AI at Infosys

Pulak Senapati, is the Senior Technology Architect – Data and AI at Infosys

Vijay Prassana is the AWS WW Partner Solution Architect for Infosys.

Copyright © 2021 IDG Communications, Inc.

Read More

Author: admin

Leave a Reply

Your email address will not be published. Required fields are marked *