Senior Security Researcher (Red Team)

You have genuine hands-on experience attacking AI systems, not just reading about them, and you enjoy breaking assumptions that others consider safe.

You continuously look for automation and AI-powered efficiencies in offensive security workflows.

As a Pindropper, you join a rapidly growing company making technology more human with the power of voice. You will work alongside some of the best and brightest. We’re a passionate group committed to excellence - but that doesn’t stop us from enjoying the journey as a team with chess and poker tournaments, catered lunches and happy hours, wellness programming, and more. Because we take our jobs seriously, we add in time for rest with Unlimited PTO.

This is a chance to shape how Pindrop attacks and defends modern AI-enabled systems. You will work at the intersection of offensive security, GenAI, deepfake defense, and cloud security, with direct impact on how we protect voice identity and high-trust customer interactions.

As a part of Pindrop, you’ll have a direct impact on our growing list of products and the future of security in the voice-driven economy. We hire great people and take care of them. Here’s a snapshot of the benefits we offer:

Competitive compensation package, including RSUs (Restricted Stock Units) for all employees, so everyone shares in our long-term success.

Remote-first environment - giving you flexibility and autonomy in how you structure your day.

As a Security Engineer (Red Team), you will help Pindrop proactively identify and exploit weaknesses across product, cloud, and AI-powered systems so we can strengthen defenses before adversaries do.

This role blends hands-on offensive security, GenAI attack simulation, security engineering, and operational partnership with blue-team, product, and AI/ML stakeholders.

Design and execute red team operations against Pindrop’s GenAI systems, LLM pipelines, RAG architectures, autonomous agents, APIs, SaaS products, and cloud environments, simulating real-world attacks across both traditional and AI-specific attack surfaces.

Conduct adversarial testing focused on prompt injection, indirect prompt attacks, jailbreaking, model extraction, training-data poisoning, data leakage, inference abuse, and unauthorized output manipulation.

Use deepfake generation, voice synthesis, and related spoofing techniques to test and attempt to defeat Pindrop’s voice authentication and deepfake detection capabilities, helping identify model robustness and detection gaps.

Develop novel attack chains that combine GenAI vulnerabilities with infrastructure, application, identity, and API weaknesses to create realistic end-to-end threat scenarios.