QA Engineer

About Lynx

AI agents are showing up in enterprise infrastructure faster than platform teams can identify and security teams can govern them. Lynx is our answer: a new product built to detect agents at runtime, understand their behaviour, and distinguish legitimate activity from misbehaviour.

We're a focused engineering team working on the hard parts, and we ship fast. Quality is what lets us keep shipping fast without breaking the trust our customers place in a security product.

About the Role

We're looking for a QA Engineer to own quality for Lynx end to end. You'll be embedded with a small, high-autonomy engineering team, building the test strategy, automation, and release confidence for a product that runs at the intersection of Kubernetes, runtime telemetry, and AI.

This is a hands-on role for someone who treats quality as engineering, not gatekeeping. You'll write automated tests, build the harnesses and fixtures the team relies on, hunt down the failures that only surface at scale, and be the person who can say with evidence whether a release is ready. You will advocate for the customer and help release great products fast, with high confidence across a range of supported platforms. Because Lynx makes security decisions from behavioural signals, you'll also help us reason about a harder question than "does it pass": is a detection correct, and how do we prove it stays correct as the models and data evolve?

You Will

• Own the test strategy for Lynx across unit, integration, end-to-end, and non-functional testing.
• Design and build automated test suites and CI pipelines that give the team fast, trustworthy signal on every change.
• Test distributed, Kubernetes-native systems: multi-node clusters, telemetry pipelines, and services under realistic load and failure conditions.
• Develop approaches for validating ML-driven behaviour — detections, risk scores, and behavioural baselines — including regression testing, ground-truth datasets, and evaluation of false-positive/false-negative rates.
• Build tooling, fixtures, and synthetic data that let the team reproduce agent behaviour and edge cases reliably.
• Investigate, isolate, and clearly report defects, driving them to resolution with engineering.
• Serve as the quality voice in design and release decisions, and help define what "ready to ship" means for a security product.

You Have

• 5+ years in software QA or test engineering, with a strong bias toward automation over manual testing.
• Proficiency in at least one programming language used for test automation (Python preferred).
• Hands-on experience testing distributed systems, APIs, and backend services.
• Working knowledge of containers and technology like Docker
• Experience building and maintaining CI/CD pipelines and integrating automated tests into them.
• Strong debugging instincts and the ability to isolate root cause across a multi-service system
• Excellent written and verbal communication — clear bug reports, clear test plans, clear recommendations.
• Must be comfortable with, and an advocate for, AI tooling in your day-to-day workflows.

Nice to Have

• Experience testing AI-powered applications or AI/ML systems is a huge advantage! Including prompt and agent testing, LLM evaluation, RAG validation, hallucination and safety testing, regression testing of AI behaviour, benchmark creation etc
• Working knowledge of Kubernetes and comfort operating clusters for testing.
• Familiarity with observability and telemetry pipelines (metrics, logs, traces; tools such as ClickHouse, Prometheus, or similar).
• Experience validating ML or data-driven systems, or building evaluation/ground-truth datasets.
• Performance, load, or chaos testing experience.
• Exposure to eBPF, kernel telemetry, or runtime security tooling.
• Open-source contributions to testing frameworks or infrastructure projects.

Work Environment

• Small team with high autonomy, working closely with engineering and reporting into the Lynx engineering leadership.
• In-office twice weekly (Mondays and Wednesdays).
• Use of Claude Code internally is expected, we build with AI, and we test the things we build with it.
• Opportunities to shape quality practice for a brand-new product from the ground up.
• Multiple office locations globally.