# TQINE0127\\_6144 - Automation : Selenium with Python**Job Type:** Contract**Work Mode:** RemoteWe are looking for a technically strong system test engineer who can validate a cloud-native inference server deployed on AWS. The role combines system testing, integration/API testing, smoke testing, automated testing, and validation activities in regulated environments. The role requires the ability to understand the architecture sufficiently to identify whether failures originate from the applications, integration, deployment, configurations, model serving layer, Kubernetes, AWS infrastructure, or integrated systems. Testing Context and Responsibilities: The inference server integrates with HALO AP as part of the broader digital pathology workflow. The engineer is responsible for independently validating that each inference server release has been correctly deployed, configured, integrated, and is fit for use across the regulated environments as specified in the approved functional and user requirement specifications. The engineer is not responsible for developing or transferring the business applications or AI models. Instead, their responsibility lies in verifying that the delivered AI models and applications work correctly once deployed into the target inference infrastructure and integrated into the wider laboratory workflow. The candidate should have hands-on system testing experience and should be comfortable in testing both individual components within the Inference server and the complete end-to end workflow. In particular, the candidate should be able to: • Perform end-to-end system testing of the integrated components with the inference server and perform smoke verification. • Execute test suite delivered by the development team of the AI models, investigate failures, and determine whether issues originate from deployment, configurations, infrastructure integration, or the applications themselves. The engineer should coordinate with the development team to determine whether additional IT-owned automated tests are required. • The engineer is not expected to act as an infrastructure engineer but must possess sufficient cloud knowledge to investigate system test failures. Working knowledge should include: o Triton/KServe model serving layer o S3 models/config/results storage o DynamoDB lifecycle o EKS/Kubernetes workloads o CloudWatch logging and monitoring • Validate delivered artifacts, including correct: o Application image container versions/digests o Model versions o Configuration versions o Environment and access configurations • Verify basic infrastructure health after deployment: o Pods running correctly o CPU/GPU nodes availability o GPU visible to workloads o Autoscaling behavior o S3 access o Logging and metrics • Test HALO AP and Inference Server integration to confirm that inference returned results behave according to the approved functional and user requirements. • The engineer should be capable of developing, maintaining, debugging, and executing automated system and integration tests • The engineer should understand and test GraphQL queries • Perform positive, negative, error-handling and recovery scenarios • Verify operational behavior such as failed jobs, retries, status handling, logging, and alarms • Perform regression testing • Verify rollback where required • Familiarity with CI/CD setup and configuration principles • Raise, report, and manage defects Required Competency The candidate should understand: • System testing • Integration testing • API testing • Automated testing (Python and Pytest) • GraphQL testing • UI testing (Selenium or equivalent framework for automated UI testing) • Regression and smoke testing • Negative and error-path testing • Test design from requirements and acceptance criteria • Root-cause analysis to locate defects • AWS fundamentals • Kubernetes/EKS • Docker containers • Terraform and Helm • CI/CD and release concepts • Container registries such as ECR • Object storage such as s3 • CloudWatch logs/metrics The candidate should provide an independent IT-side verification that the deployed infrastructure as code release and deployed applications/AI models work as intended at system and end-to-end level, and that sufficient evidence exists to accept the candidate release in the broader laboratory governed environments.