New opportunity

Lead Cybersecurity Engineer- AI Red Team

Societe Generale · Bangalore, India

About this role

Societe Generale is looking for a Lead Cybersecurity Engineer for their AI Red Team in Bangalore, India, to build production-grade GenAI capabilities for cybersecurity use cases. The role focuses on designing and delivering LLM and RAG solutions while leveraging cybersecurity and Red Team exposure to ensure practical, secure, and real-world security operations.

Responsibilities

Design, build, and industrialize GenAI solutions using LLMs, RAG, embeddings, vector search, agents, APIs, and orchestration frameworks.
Own platform engineering aspects including ingestion pipelines, observability, evaluation, CI/CD, monitoring, security controls, and operational readiness.
Partner with cybersecurity stakeholders to translate threat analysis, investigation, and security automation needs into AI-enabled workflows.
Apply working knowledge of AI security risks such as prompt injection, data leakage, hallucination, model misuse, and unsafe tool usage.
Collaborate with Red Team or offensive security teams to support use-case design, validation, and practical adoption.

Required qualifications

8+ years overall experience with strong hands-on AI/GenAI engineering and platform delivery experience.
Strong hands-on GenAI engineering experience with LLMs, RAG, embeddings, vector databases, prompt engineering, evaluation, and production deployment.
Ability to design scalable AI systems covering APIs, orchestration, pipelines, monitoring, observability, CI/CD, and operational controls.
Strong Python or backend engineering skills, secure software development, microservices, cloud-native design, and integration with enterprise systems.
Working understanding of cybersecurity concepts such as threat analysis, vulnerability management, identity, cloud security, security telemetry, and risk controls.

Preferred qualifications

Exposure to AI security topics such as prompt injection, jailbreaks, data leakage, model abuse, hallucination, and insecure agentic workflows.
Familiarity with Red Team, penetration testing, adversary simulation, MITRE ATT&CK, or purple team practices.
Experience integrating AI capabilities with SIEM, SOAR, EDR, threat intelligence, vulnerability management, or cloud security platforms.
Experience working in regulated enterprise environments with governance, auditability, and responsible AI expectations.

For full details about this role, click “Apply on company site” to view the employer’s original posting.

Skills for this role

GenAILLMRAGEmbeddingsVector SearchAI AgentsPythonPrompt EngineeringCI/CDCloud SecurityPenetration TestingSIEMSOAREDR

Checking your profile…

Preparation suggestions, not the employer’s actual interview questions. Always check the original posting for current requirements.

Understand the role

What matters most

Design and Delivery of Production-Grade GenAI Solutions
Build and industrialize scalable generative AI applications using LLMs, RAG, vector databases, embeddings, and orchestration frameworks to support real-world security use cases.
Platform Engineering and Operational Readiness
Own end-to-end platform aspects including data ingestion pipelines, monitoring, evaluation, observability, security controls, and robust CI/CD pipelines.
Collaboration with Cybersecurity and Red Teams
Partner with security stakeholders and offensive security teams to translate threat analysis and security automation needs into practical, validated AI-enabled workflows.
Mitigation of AI Security Risks
Apply working knowledge of common generative AI vulnerabilities such as prompt injection, data leakage, hallucination, model misuse, and insecure agentic workflows.

Likely challenges to think through

Balancing Advanced GenAI Architecture with Core Security Constraints
The challenge lies in rapidly delivering complex LLM/RAG pipelines while rigorously applying security controls against threats like prompt injection and data leakage. Prepare by reviewing secure software development practices and AI safety frameworks.
Bridging Deep Engineering with Offensive Security Domain Context
While deep Red Team expertise is not mandatory, the role requires effective communication with offensive security teams and understanding threat intelligence or vulnerability management. Prepare by studying penetration testing concepts and adversary simulation frameworks like MITRE ATT&CK.

Practice your answers

Technical practice

Designing Production-Grade RAG Systems

What to coverExplain how you would design a production-grade Retrieval-Augmented Generation (RAG) pipeline for cybersecurity use cases. Discuss how you handle document ingestion, embedding generation, vector database selection, and retrieval latency in enterprise environments.

Mitigating Prompt Injection and Model Abuse

What to coverWhat strategies and security controls do you implement to prevent prompt injection, jailbreaks, and unsafe tool usage in LLM-powered applications? Address how input sanitization and output validation can safeguard autonomous agents.

Platform Engineering and CI/CD for AI Systems

What to coverHow do you establish robust CI/CD pipelines, monitoring, and observability for generative AI solutions? Discuss how you track model drift, evaluation metrics, and operational readiness in production deployments.

Integrating AI Capabilities with Security Operations

What to coverHow would you translate threat analysis, vulnerability management, and incident investigation workflows into AI-enabled capabilities? Discuss integration strategies with security telemetry, SIEM, SOAR, or EDR platforms.

Secure Microservices and Cloud-Native Python Engineering

What to coverDescribe how you design secure, scalable microservices in Python for enterprise system integration. Focus on secure software development principles, API design, and cloud-native architecture controls.

Collaborating with Red Teams and Adversary Simulation

What to coverHow can an AI engineering team collaborate effectively with Red Teams or offensive security units using frameworks like MITRE ATT&CK? Discuss how penetration testing insights inform the validation of GenAI use cases.

Handling Data Leakage and Confidentiality in LLMs

What to coverWhat techniques do you use to prevent sensitive corporate or customer data leakage through LLM completions and embeddings? Address data anonymization, access controls, and boundary enforcement.

Evaluating Agentic Workflows and Tool Safety

What to coverHow do you evaluate and test autonomous AI agents that have access to external tools and enterprise APIs? Discuss methods for detecting model misuse and ensuring safe execution boundaries.

Behavioural practice

Translating Complex Security Needs into AI Solutions

What to coverDescribe a situation where you partnered with cybersecurity or technical stakeholders to understand complex operational needs. Explain how you translated those requirements into a practical, scalable technical solution, detailing the situation, your actions, and the outcome.

Balancing Speed of Delivery with Rigorous Security Controls

What to coverShare an experience where you had to balance rapid platform delivery or prototyping with strict security and governance requirements. Discuss the trade-offs you evaluated, the actions you took, and how you ensured operational readiness.

Collaborating with Cross-Functional Offensive Security Teams

What to coverTell us about a time you worked alongside Red Teams, penetration testers, or domain specialists whose expertise differed from your own. How did you bridge communication gaps, incorporate their feedback, and validate use-case adoption?

Managing Production Incidents and System Observability

What to coverProvide an example of a challenging production issue or unexpected system failure in an AI or data platform you managed. Outline how you utilized monitoring, logs, and observability tools to diagnose and resolve the problem under pressure.

Sharpen your knowledge

Topics to prepare

Generative AI and RAG Architecture
Review core concepts of Large Language Models, Retrieval-Augmented Generation, embeddings, and vector databases. Exercise: Design a production-grade RAG pipeline architecture diagram incorporating document ingestion, embedding generation, and vector search. Relevance: Directly aligns with the primary responsibility to build production-grade GenAI capabilities for cybersecurity workflows.
AI Security Risks and Threat Modeling
Study vulnerabilities specific to AI systems, such as prompt injection, jailbreaks, data leakage, hallucination, and unsafe agentic workflows. Exercise: Conduct a threat model on an LLM-powered assistant integrated with enterprise tools to identify potential prompt injection and model misuse vectors. Relevance: Fulfills the requirement to apply working knowledge of AI security risks and collaborate with red team workflows.
Secure Platform Engineering and CI/CD
Understand scalable AI system design, cloud-native architecture, observability, monitoring, and operational controls. Exercise: Outline an automated CI/CD and evaluation pipeline for monitoring LLM drift and performance metrics in production. Relevance: Essential for owning platform engineering aspects, including ingestion pipelines and operational readiness.

Skills to revise

Python and Backend API Integration
Practice building secure microservices and robust API integrations using Python. Exercise: Develop a Python-based microservice endpoint that securely interfaces with an LLM orchestration framework while implementing input sanitization and error handling. Relevance: Directly supports the core requirement for strong Python and backend engineering skills in enterprise systems.
Vector Databases and Orchestration Frameworks
Gain hands-on proficiency with vector search tools and LLM orchestration frameworks. Exercise: Set up a local vector database index and write code to query embeddings using an orchestration framework. Relevance: Directly tied to the requirement to design, build, and industrialize GenAI solutions using vector search and orchestration tools.
Security Telemetry and SIEM/SOAR Integration
Familiarize yourself with security event monitoring and automation platforms. Exercise: Map out how AI-generated insights and threat intelligence feeds can be integrated into SIEM or SOAR platforms for automated security workflows. Relevance: Aligns with preferred qualifications for integrating AI capabilities with security telemetry and enterprise platforms.

Your preparation checklist

Prepare a GenAI Platform Architecture Walkthrough
Compile a detailed system architecture diagram and documentation showing your hands-on experience designing and industrializing LLM, RAG, and vector search applications with robust ingestion pipelines and orchestration frameworks.
Draft Evidence of Secure Python Backend Delivery
Gather concrete examples and code snippets from your past work demonstrating strong Python skills, secure microservices, cloud-native design, and enterprise system integration over your 8+ years of experience.
Build a Threat Model for AI Security Risks
Prepare a structured case study or analysis addressing common generative AI vulnerabilities—such as prompt injection, data leakage, and model misuse—and how you implement security controls and guardrails.
Outline Integration Strategies with Security Operations
Document examples of how you have partnered with cybersecurity stakeholders or integrated AI capabilities with telemetry, threat intelligence, and monitoring platforms like SIEM or SOAR.
Formulate Thoughtful Questions on Red Team Workflows
Prepare targeted questions for the interview team regarding how the AI engineering group collaborates with offensive security and Red Teams to validate use cases in a regulated enterprise environment.

Availability last checked: 4 Nov 2026. Status can change after a check; confirm on the employer’s website. Report an issue · Our checking policy

KEEP EXPLORING

Similar AI jobs

Related skills and specializations in India. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…