Staff SDET, AI Gateway

Please note, this team is hiring across all levels and candidates are individually assessed and appropriately leveled based upon their skills and experience.

About the Role

Netskope AI Gateway helps global enterprises securely adopt and scale Generative AI. We are looking for a Staff Software Development Engineer in Test (SDET) to own quality for significant areas of the AI Gateway product, its supporting services, and its APIs, across SaaS and on-prem/hybrid customer environments.

This is a hands-on individual-contributor role with real ownership. The areas assigned to you are yours: you set the testing direction and build what is needed to execute it, bringing hard calls to the QE lead as a peer rather than waiting for a specification or a checklist before you act.

Depth matters more than volume in this role. We are looking for someone who can get a long way into an ambiguous problem before needing help, go deep when a failure crosses service, infrastructure, and network boundaries, and work directly with developers and product managers to get the underlying problem fixed rather than routed around.

What You Will Be Doing

• Own Quality for Your Areas: For the features and services assigned to you, own the outcome: reliability, correctness, performance, security-related behaviors, and user experience expectations. Decide what needs testing and at which layer, set the bar for shipping that area, and be accountable for what gets through.
• Turn Ambiguity into a Plan: Start from partial requirements, an architecture diagram, and a conversation with the developer, and produce a concrete test approach. Surface the risks nobody has written down yet, decide which ones matter, and state clearly what you are choosing not to cover and why.
• Build the Automation: Design and implement the automated suites and the frameworks, harnesses, and fixtures behind them. Write them as real software: readable, maintainable, and quick to triage when they fail. Keep flakiness low through deterministic test design, environment isolation, and disciplined use of mocking and service virtualization.
• Use AI Where It Earns Its Place: Bring AI and agent-based tooling into the testing workflow: generating scaffolding and test data, navigating unfamiliar code, summarizing logs and failures, and shortening the path from symptom to hypothesis. Be equally clear about the limits. AI-generated tests can look thorough while asserting nothing that matters, so you check what the assertions actually prove. The tool can draft the work, but the coverage decision and the verdict stay yours.
• Validate the Gateway Data Path: Verify that AI Gateway stays compatible with the provider APIs it fronts, that policy matches exactly the requests it should, and that the resulting action and logging are correct.
• Debug Deep and Across Boundaries: Be the person who gets to the bottom of hard failures. Trace problems across microservices, asynchronous workflows, Kubernetes networking, TLS and proxy behavior, and CI infrastructure, and come back with a root cause rather than a reassignment.
• Keep CI/CD Trustworthy: Integrate your tests into CI/CD with attention to signal-to-noise, parallelization, execution time, and reporting that points at the actual failure. Diagnose pipeline and environment problems instead of retrying around them.
• Share Production Ownership: Own the production service alongside the rest of the AIG engineering team. Take part in the shared on-call rotation, monitor production health, and troubleshoot and resolve critical production issues.
• Push Problems Upstream: Work directly with developers and product managers. Raise testability and instrumentation needs while designs are still open, argue for the fixes that prevent whole classes of defects, and make sure recurring failure patterns change how the next feature is built. Write bug reports a developer can act on immediately: reproduction steps, logs, traces, and a clear statement of impact.

What Success Looks Like

• The areas you own ship with high confidence, and the defects that do escape are rare and understood.
• A vague problem handed to you comes back as a clear plan, a working suite, and an explicit list of the risks you accepted.
• Hard cross-boundary failures land with you and leave with a root cause.
• Your automation is trusted: when it fails, people believe something is actually broken.
• AI makes the work faster, and the coverage still holds up when someone reviews it.
• Developers change designs because of risks you raised early, and the same defect does not come back twice.

Required Skills and Experience

• Experience: 8+ years in software quality engineering / SDET / test automation, including 5+ years on distributed systems, cloud services, or security/networking products. You have owned quality for a substantial product area with little supervision.
• Autonomy: Track record of taking an under-specified problem and driving it to a working solution, including deciding what not to do. Comfortable being the only person working on something.
• Programming: Strong software engineering skills in Python (preferred) or Golang. You write automation, libraries, and tooling that hold up as production software rather than as scripts.
• AI-Assisted Engineering: Practical experience using LLM-based tools or coding agents in real engineering work, and a clear view of where they help and where they mislead. You can point to a case where you caught what the AI missed.
• Debugging Depth: Demonstrated ability to isolate root cause in complex failures spanning services, infrastructure, and the network layer, rather than escalating them onward.
• API and Distributed Systems Testing: Extensive hands-on experience testing REST APIs, microservices, and asynchronous workflows.
• Test Design: Deep command of test strategy, risk-based testing, and layered test design; able to make and defend explicit trade-offs on coverage, cost, execution time, and confidence.
• CI/CD: Proven ability to build and troubleshoot CI/CD pipelines and the infrastructure that keeps large test suites reliable.
• Containers and Kubernetes: Strong experience running and testing services on Docker/Kubernetes, and debugging failures involving networking, service discovery, and configuration.
• Networking: Solid working knowledge of HTTP/S, TLS, proxy/gateway concepts, DNS, and the failure modes that affect AI Gateway deployments.
• Defect Management: Excellent defect isolation, triage, and tracking discipline (e.g., Jira), with follow-through until the systemic fix lands.
• Working with Engineering and Product: Able to hold a technical discussion with developers as a peer and to explain risk and trade-offs to product managers clearly. Exceptional verbal and written English for complex technical collaboration.

Nice to Have

• GenAI Ecosystem: Familiarity with the request and response schemas of major LLM provider APIs, such as OpenAI, Anthropic, Google, or Bedrock, including their streaming formats and how these APIs change over time.
• Cloud Platforms: Experience with AWS, Azure, or GCP, and validating hybrid/on-prem customer deployments.
• Cybersecurity: Knowledge of secure web development, ZTNA, DLP, and proxy/gateway product concepts.
• Performance and Scale: Experience designing load and performance tests and analyzing latency/throughput for high-traffic APIs and gateways.
• Resilience Testing: Fault injection, chaos, or failure-mode testing of distributed services.
• Observability: Hands-on use of metrics/logging/tracing tooling (e.g., Prometheus, Grafana, OpenTelemetry) for faster test and defect triage.

Education

BS in Computer Science or equivalent degree required ; MS in Computer Science or equivalent is strongly preferred.

#LI-SK3