Senior Staff AI Engineer, (AI Quality & Test Automation)

What if the work you did every day could impact the lives of people you know? Or all of humanity? At Illumina, we are expanding access to genomic technology to realize health equity for billions of people around the world. Our efforts enable life-changing discoveries that are transforming human health through the early detection and diagnosis of diseases and new treatment options for patients. Working at Illumina means being part of something bigger than yourself. Every person, in every role, has the opportunity to make a difference. Surrounded by extraordinary people, inspiring leaders, and world changing projects, you will do more and become more than you ever thought possible. Location San Diego, CA. Not remote. Position Summary We are hiring a Senior Staff AI Engineer to set the technical direction for quality engineering across Illumina's enterprise platforms, with AI agents doing a growing share of the testing work. You will be the most senior individual contributor in QA: you design the test architecture, lead our Tricentis platform, and make agentic testing safe enough for regulated, GxP-relevant systems. Technical leadership of 10-20 QA and automation engineers across vendors and teams. The role sits in Enterprise AI and partners with Data Engineering, application teams (SAP, Salesforce, ServiceNow, Workday, M365) and Quality & Regulatory. It covers two equally important missions: using AI agents to test our software faster, and testing the AI agents we build so they can be trusted in production. Responsibilities Own the enterprise test strategy. Define a risk-based, shift-left and shift-right strategy across SAP S/4HANA, Salesforce, ServiceNow, Workday, custom apps, data pipelines and AI agents. Lead the Tricentis platform. Architect Tosca, qTest, NeoLoad, SeaLights, LiveCompare and Tricentis AI Workspace as one quality platform, including standards, reuse libraries, licensing and roadmap. Build agentic testing at scale. Design and run multi-agent test workflows that generate, execute, heal and triage tests, with clear human approval points and audit trails. Test our AI agents. Create the evaluation framework that gates every agent release: golden datasets, LLM-as-judge scoring, red teaming, tool-call correctness, safety and cost checks. Connect agents to tools through MCP. Integrate Tricentis remote MCP servers and our own MCP servers with Claude, GitHub Copilot, Cursor and Codex under the Enterprise AI gateway and policy. Embed quality in CI/CD. Make test impact analysis, quality gates and release risk scoring standard in every pipeline, so teams ship on evidence rather than gut feel. Deliver validated quality for GxP. Partner with Quality & Regulatory to apply FDA CSA and GAMP 5 (2nd ed.) so automated and AI-assisted testing produces Part 11 compliant evidence. Raise the bar for engineers. Mentor senior and staff engineers, run design reviews, write the playbooks, and lead the QA community of practice. Measure and report value. Own the quality scorecard (escape rate, automation coverage, cycle time, cost per test) and report it to engineering and business leaders. Experience & Education 12+ years in software quality engineering, with 5+ years as a lead, principal or staff level engineer setting strategy across multiple teams. 5+ years of hands-on Tricentis Tosca experience at enterprise scale, plus working knowledge of qTest and at least one of NeoLoad or SeaLights. Proven delivery of test automation for packaged enterprise apps, ideally SAP S/4HANA or ECC and Salesforce. 1+ year building or operating LLM-based or agentic testing in production, not only proofs of concept. Strong coding in at least two of Python, TypeScript/JavaScript, Java or C#. Deep CI/CD experience with GitHub Actions, Azure DevOps, GitLab or Jenkins, including quality gates. Experience testing in regulated environments (GxP, medical device, SOX or similar) and producing audit-ready evidence. Clear written and spoken communication with engineers, product owners and executives. Bachelor's degree in Computer Science, Engineering or a related field, or equivalent experience. Preferred qualifications Tricentis certifications (Tosca Architect, qTest, NeoLoad); ISTQB Advanced or Expert, or ISTQB Certified Tester AI Testing (CT-AI). AWS Certified Machine Learning or AI Practitioner, or Databricks Generative AI Engineer certification. Published MCP servers, eval suites or open source test tooling. Experience in life sciences, genomics, diagnostics or medical devices. Experience leading an SAP S/4HANA migration or RISE program test effort. Tricentis platform skills We expect expert depth in Tosca and qTest, and practical command of Tricentis's agentic products released in 2025 and 2026. Tosca (Commander, Cloud, Vision AI, model-based test automation): Architect module and test case design, reusable libraries, test data service, distributed execution and Tosca Cloud migration Tosca Copilot and Agentic Test Automation: Generate full test cases from natural language and Vision AI, review output, and govern what agents may change qTest (Manager, Insights, Explorer, qTest Copilot, Agentic Test Creation): Run test management, traceability to requirements and Jira, and agent-assisted test authoring Tricentis AI Workspace: Onboard and orchestrate Tricentis and third-party agents with approvals, governance and human oversight Tricentis remote MCP servers: Connect Tosca, qTest and NeoLoad to Claude, Copilot, Cursor and our own agents through the Enterprise AI MCP gateway SeaLights (quality intelligence): Apply test impact analysis and code coverage to cut regression runs and find untested change NeoLoad and Agentic Performance Engineering: Design performance and load tests, including AI-generated user paths for SAP and web apps LiveCompare: Run SAP change impact analysis to scope regression for transports and upgrades Testim and Tricentis Device Cloud: Automate web and mobile UI tests with self-healing locators across real devices Tricentis Data Integrity: Validate data pipelines, migrations and reports end to end Newer Tricentis Labs tools (Aida, AgentScore, Release Risk Intelligence): Pilot autonomous exploration, agent quality scoring and release risk scoring, and advise on adoption Tricentis certifications (Tosca Automation Specialist and Architect, qTest, NeoLoad) are strongly preferred. Product names reflect Tricentis's portfolio as of October 2026 ( Pulse 2.0 interview , VKTR , Business Wire ). AI agentic testing skills You will be fluent on both sides: agents that test software, and testing of the agents themselves. Agents that test software Designing multi-agent QA workflows (planner, generator, executor, healer, triage, reporter) with defined roles, hand-offs and stop conditions. Building agents with Claude Agent SDK, Amazon Bedrock AgentCore, LangGraph, CrewAI, Microsoft Copilot Studio or Databricks Agent Bricks. Model Context Protocol (MCP) servers and clients, and Agent2Agent (A2A) protocol for agent-to-agent hand-offs. AI-native browser and UI automation: Playwright MCP, Claude computer use, browser agents and visual AI (Tricentis Vision AI, Applitools). Agentic coding assistants for test code: Claude Code, GitHub Copilot agent mode, Cursor and Codex, with review and branch protection rules. Autonomous exploratory testing, self-healing locators, and AI-driven defect triage, clustering and root cause suggestions. Prompt and context engineering: skills, system prompts, retrieval of requirements and past defects, structured outputs. Synthetic test data generation with privacy controls (no real PHI or PII in prompts). Testing AI agents and LLM applications Evaluation design: golden datasets, rubric scoring, LLM-as-judge with calibration against human labels, pass@k and regression baselines. Eval tooling such as Bedrock AgentCore Evaluations, MLflow 3 GenAI evaluation, DeepEval, Ragas, Promptfoo, LangSmith, Arize Phoenix or Braintrust. Agent-specific check