Website:
korelex.ai
Job details:
Company Description Korelex.ai specializes in making AI systems provable and trustworthy for enterprises that adopt AI faster than they can properly evaluate it. The company focuses on red-teaming production AI systems—covering models, RAG pipelines, agents, and the underlying APIs, tools, and retrieval layers—to expose what a determined attacker could find. Korelex.ai delivers findings mapped to real-world security and compliance frameworks, providing signed, dated artifacts that customers can share with auditors, boards, and clients. As an independent organization that does not build or sell the systems it tests, Korelex.ai avoids conflicts of interest and emphasizes accurate reporting of both controls that fail and controls that hold under attack. The team also contributes to the AI security ecosystem through resources such as the Daily Briefing on AI Security and tools like DPDP assessment.
Role Description: As a Red Team Researcher at Korelex.ai, you will design and execute adversarial tests against production AI systems, including large language models, RAG pipelines, agents, and their supporting infrastructure. You will investigate vulnerabilities, develop attack methodologies, and document findings in a way that aligns with industry security and compliance frameworks. Your day-to-day work will include creating realistic attack scenarios, collaborating with engineers and product teams to understand system architecture, and producing clear, actionable reports that distinguish confirmed issues from open questions and resilient controls. You will also contribute to internal tooling, methodologies, and research to advance AI security practices and help refine Korelex.ai’s red-teaming processes.
This is a full-time, on-site role based in Bengaluru.
Qualifications
- Strong foundation in security research and red teaming, with experience analyzing complex systems, identifying vulnerabilities, and simulating adversarial behavior against AI or software platforms.
- Practical knowledge of modern AI systems, including LLMs, RAG pipelines, agents, and APIs, with the ability to reason about data flows, retrieval layers, and multi-tenant architectures.
- Solid understanding of application security and secure design principles, such as access control, data privacy, threat modeling, and common attack vectors relevant to AI-driven products.
- Proficiency with scripting or programming languages (e.g., Python, JavaScript) to build custom tools, automate experiments, and analyze large volumes of test data.
- Ability to produce clear, structured technical documentation and reports that map findings to recognized security and compliance frameworks and distinguish between confirmed, observed, and open issues.
- Familiarity with industry standards and regulations relevant to AI and data protection (e.g., DPDP, GDPR, SOC 2, ISO 27001) is highly beneficial.
- Comfort working closely with engineering and product teams
Click on Apply to know more.
