What challenges did Cognizant face before AI implementation?

Contact centers faced high call volume and staffing constraints. Agent experiences were inconsistent across channels and teams. Agents had knowledge gaps and limited real-time support during live calls. Post-call analytics for coaching and quality assurance were weak.

Source

ExpandedEvidence: Low25/100

Agent AI Ally: Real-time agent assist and post-call analytics using Amazon Connect and Amazon Bedrock (Cognizant)

Use case typeCustomer service automationUpdated Jun 13, 2026

Cognizant’s Agent AI Ally is a contact center solution for live agent support, virtual training, and post-call analytics. The solution uses Amazon Connect for call routing and an agent workspace, Amazon Bedrock with Anthropic Claude for generative AI suggestions, Amazon Kinesis Video Streams for real-time audio capture, Amazon Transcribe for speech-to-text, and AWS Lambda for transcription and analytics processing.

Organization: Cognizant
Industry: Professional Services
Location: United States
Published: March 2025

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: Cognizant
Provider: AWS
Maturity: Unknown
Linked source: AWS Partner Network

No explicit deployment-stage evidence found.

Customer identity supportedSource describes one deploymentMaturity evidence evaluated

Primary read

Use case focus

Showing 3 of 5

1Agent assist
2Contact center analytics
3Conversation intelligence

Calls are routed to agents in Amazon Connect.
Real-time call audio is captured with Amazon Kinesis Video Streams and transcribed with Amazon Transcribe.
An AI assistant interface uses Amazon Bedrock to suggest follow-up questions and solutions during the call.
After calls, AWS Lambda supports analytics that generate summaries, customer sentiment, and call categorization for coaching and performance improvement.
The solution also includes virtual training with simulated customer calls.

Technologies

Amazon Connect Amazon Bedrock Anthropic Claude 3 Amazon Kinesis Video Streams Amazon Transcribe AWS Lambda

Architecture

When a customer calls, Amazon Connect routes the call to an available agent. The agent receives the call in the Amazon Connect workspace and simultaneously accesses the Agent AI Ally interface. Amazon Kinesis Video Streams captures the call audio in real time; Amazon Transcribe converts speech to text; Amazon Bedrock’s Anthropic Claude model suggests follow-up questions and solutions during the conversation. After the call, AWS Lambda analyzes the recording and produces post-call analytics in seconds.

Sources & evidence1

Evidence: Low25/100Evidence strength