Principal Software Engineer, AI-Driven Test Automation

Roca Alliances; Medellín; Sur place; COP 360 000 000 - 480 000 000; Plein temps

Résumé du poste

Roca Alliances is seeking a Principal Software Engineer in Test to own product quality for a global SaaS field service platform. You will define test strategy, automation architecture, and AI agents to execute it.

This role is a deeply technical IC, closer to platform engineering than QA, with no people-management duties. You will own the quality signal, automate critical customer workflows end-to-end, and drive a release gate based on realistic data.

Qualifications

• Deep experience designing test automation for a large, long-lived web application.
• Hands-on experience with LLMs, agent loops, and evaluation harnesses in CI.
• Strong SQL and relational data skills for test data in complex schemas.
• CI/CD ownership: pipelines, gating, flake management and run-time budgets.
• Advanced English.
• Experience with AI-driven testing approaches is a plus.

Responsabilités

• Own test strategy across web, mobile, and API layers and data.
• Design and operate automation architecture and data strategy for testing.
• Develop AI agents that derive test cases from tickets and code changes and review PRs.
• Maintain engineering practices for quality, including standards and guardrails.
• Deliver a clear quality signal to engineering and executives.

Connaissances

Test automation design

LLM agent loops

SQL

CI/CD ownership

English proficiency

AI in testing

Outils

Jenkins

Bitbucket

Jira

Confluence

Terraform

Datadog

Grafana

Description du poste

Roca Alliances is seeking a Principal Software Engineer in Test to own product quality for a global SaaS field service platform. You will define test strategy, automation architecture, and AI agents to execute it.

This role is a deeply technical IC, closer to platform engineering than QA, with no people-management duties. You will own the quality signal, automate critical customer workflows end-to-end, and drive a release gate based on realistic data.