TCS built an agentic test orchestrator with Amazon Bedrock AgentCore to automate testing in the BFSI sector
TCS has built an agentic test orchestrator on Amazon Bedrock AgentCore to automate testing of BFSI applications – specialized agents, API and UI tests, tenant data isolation in micro-VMs.
Tata Consultancy Services (TCS) has built an agentic test orchestrator based on Amazon Bedrock AgentCore, designed to automate application testing for clients in the BFSI sector (banking, financial services, insurance). According to TCS and AWS, who co-authored the text, the goal was not to replace testers but to eliminate the repetitive work of converting requirements into test scenarios and running them, so that engineers can focus more on risk analysis and release decisions.
The orchestrator breaks testing down into specialized agents, each with its own tools, validation rules, and model chosen according to the type of task (e.g. scenario design versus code generation). A deterministic supervisor manages the workflow, task retries, error classification, and reporting. The platform supports both API testing and web UI testing, and accepts input from Jira, business requirement documents, Swagger specifications, and GitLab repositories, from which it generates test scenarios, Gherkin files, and executable tests in Java or Python.
This is a multi-tenant SaaS solution on AWS: a shared control layer (Amazon EKS, CloudFront, WAF) serves all clients, while each tenant has its own isolated application layer – a separate Kubernetes namespace, Cognito user pool, KMS key, S3 bucket, DynamoDB tables, and Bedrock Knowledge Base. Agents run in the Amazon Bedrock AgentCore runtime, where, according to the description, each session is isolated in its own micro-VM; tests are run on Amazon ECS and results are stored in S3 and DynamoDB.
The source text is a technical case description published on the AWS Machine Learning Blog and does not describe specific speedup figures, pricing, or general availability of the platform for other companies – it is an internal TCS platform built on AWS services.
Why it matters
The case shows a concrete enterprise architecture for replacing manual creation and maintenance of test scripts with a network of agents with their own models and validation rules, while maintaining the data isolation of individual clients required in regulated industries. It is relevant for companies and QA teams considering the deployment of agentic systems into processes with an audit trail and tenant-level security, not for ordinary end users.
Relevant practical impact
What this means
For a business
For companies running large-scale testing (especially in the regulated BFSI sector), the TCS case demonstrates a concrete architecture for replacing manual writing and maintenance of test scripts with a network of specialized agents and a deterministic supervisor, while maintaining data isolation of individual clients at the micro-VM, key, and identity level – relevant for companies evaluating…
ProcessesCheck the original
Event sources
only one source so far · 1 publisher, 0 independent. We count feeds from the same owner only once.