Tester; Remote

United States

Suitability/Public Trust

Fully remote

Information Technology

GovCIO is currently hiring a Software Development Engineer in Test (SDET), Test Engineer, to join cross-functional product teams building and modernizing federal digital services on AWS Gov Cloud.

This position is in the Continental United States (CONUS) and remote.

A GovCIO SDET works in a cross-functional product team where everyone owns quality. The SDET plans testing for each story, builds and maintains automated tests, and reports what is covered and what is not. The role is hands-on, automation-first, and deadline-driven. Story testing starts and finishes inside the sprint; the SDET also creates the longer tests that span sprints, such as functional, regression, end-to-end, integration, performance, and load.

A GovCIO SDET is not a script executor, not a click-through tester, and not the last gate before release. Quality starts at refinement: the SDET makes acceptance criteria testable, works from user stories without waiting for a BRD or FRD, and audits the tests developers write. The SDET knows what to regression test and where to look when something breaks, and can evaluate non-deterministic output using measurable thresholds that separate model variance from real defects.

Designs, develops, and implements testing methods and equipment to ensure that the product will perform to specifications. Plans and arranges the labor, schedules, and equipment required for testing and evaluating the product. Designs test cases for test plans and creates test scenarios in which tests are carried out. Compiles data, defines required changes and reports defects and problems that occurred during the test process.

Implements and participates in the walkthrough of the product. Automated testing may be used for leveling based on hiring manager discretion.

• Consults with the development and/or requirements team to coordinate and test applications.
• Performs analysis, diagnostics and preliminary evaluations of products. Develop testing procedures.
• Develops test plans, scripts, scenarios, functional tests, regression tests, and deployment tests.
• Implements the testing processes and ensures all test activities follow those processes. Produces reports on each test activity and tracks test metrics.
• Monitors for continual test process improvement through failure analysis, throughput analysis, yield analysis, and report to management periodically.

• Work as a full member of a Scrum team. Take part in sprint planning, backlog refinement, and design walkthroughs. Make user stories and acceptance criteria testable. Write your own stories and tasks.
• Design and run positive and negative tests: functional, integration, regression, API, and end-to-end. Find or create the test data. Pace the work so testing finishes inside the sprint.
• Build and maintain the automated test suites. Use Cypress and/or Playwright for the front end, Postman or scripted tests for REST APIs, and Locust for load and performance. Put the suites in the CI/CD pipeline with the Dev Ops engineers. Build and manage test environments on AWS Gov Cloud.
• Use Amazon S3 to check test data and artifacts. Use Amazon Cloud Watch and Logs Insights queries for logs, metrics, and troubleshooting during test runs.
• Own regression testing for the product. Know how the system fits together. Decide what must be re-tested and in what order. Keep the regression suite current as the product changes.
• Audit the unit, contract, integration, and end-to-end tests the developers write. Work with them to improve coverage and quality.
• Check data in PostgreSQL and other data stores with SQL. Check JSON, XML, and file content with scripts.
• Test machine learning and deep learning features, especially the large language models and vision models behind OCR and intelligent document processing (IDP). Build and maintain ground-truth sets. Measure accuracy at the character, field, and document level and as a straight-through rate. Set threshold-based acceptance criteria and be able to say where each number came from. Check whether confidence scores predict correctness.
• Evaluate generative AI features. Run eval harnesses against versioned prompts. Test schema conformance and answer correctness as separate questions. Catch values the model made up that are not in the source document.
• Automate eval runs in CI so regressions show up before UAT.
• Keep the Requirements Traceability Matrix, defect records, test…