AI Solutions Security Engineer
Replicon Software · Bengaluru · onsite · Posted 9d ago
Undisclosed3–7 yrsfulltimeSourced from
About a minute. You'll make an account on the way through.
About this role
Key Responsibilities
AI Security Threat Defense
- Design and implement security architecture and controls for LLM, Generative AI, RAG, and agentic AI applications.
- Perform threat modeling and AI-specific risk assessments aligned to frameworks such as OWASP Top 10 for LLM Applications, MITRE ATLAS, and NIST AI RMF.
- Defend against prompt injection, jailbreaks, data exfiltration, model and data poisoning, insecure output handling, and excessive agent autonomy.
- Conduct AI red teaming, adversarial testing, and penetration testing of models, prompts, agents, and integrations; track findings through remediation.
- Secure MCP servers, agent-to-agent (A2A) communication, tool calling, and plugin integrations with least-privilege access, authentication, and authorization.
- Protect models, embeddings, vector stores, prompts, and training data with encryption, access controls, secrets management, and network isolation.AI Guardrails Runtime Protection
- Design, implement, and tune guardrails for input and output filtering, content safety, topic restrictions, grounding checks, and hallucination mitigation.
- Build and operate PII, sensitive data, and secrets detection and redaction across prompts, retrieved context, and model responses.
- Evaluate and integrate guardrail tooling such as AWS Bedrock Guardrails, Azure AI Content Safety, NeMo Guardrails, Guardrails AI, and LLM Guard.
- Build automated safety, security, and bias evaluation suites and run them as quality gates in CI/CD pipelines.AI Compliance, Policy Governance
- Translate AI policies, Responsible AI principles, and regulatory requirements into enforceable technical controls and policy-as-code.
- Support compliance with frameworks and regulations such as the EU AI Act, NIST AI RMF, ISO/IEC 42001, SOC 2, ISO 27001, GDPR, and CCPA.
- Establish and run AI risk review, and exception processes with security, privacy, legal, and business stakeholders.Monitoring, Detection Incident Response
- Implement logging, monitoring, and alerting for AI-specific security events such as policy violations, anomalous usage, data leakage, and model abuse.
- Integrate AI telemetry with SIEM, SOAR, and observability platforms to support detection and response.
- Develop AI incident response playbooks and participate in triage, investigation, and post-incident reviews.AI Solutions Engineering
- Collaborate with AI data scientists and solutions teams to design, build, and deploy secure Generative AI, RAG, and agentic solutions using LLMs such as GPT, Claude, Gemini, Llama, and Mistral.
- Develop production-grade Python services, APIs, and reusable secure components, such as guardrail libraries, SDKs, and templates.
- Integrate AI services with enterprise systems, identity providers, and data platforms using secure design patterns.Cloud DevSecOps
- Secure AI workloads on AWS (preferred) and Azure, including IAM, VPC and private networking, KMS, and secrets management.
- Embed security scanning in CI/CD pipelines, including SAST, DAST, dependency and SBOM checks, container and image scanning, and infrastructure-as-code scanning.
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Disclaimer: This job posting has been aggregated from external source. Role details, content, and availability are subject to change. Applicants are advised to confirm the latest information directly on the company website before applying.
Sourced from naukri
Skills
SOCUsageremediationamazon web servicesIso 27001MonitoringPythonmanagementPenetration Testingcompliance
About Replicon
Replicon Software
Software Product
771
employees
1996
30 years old
Calgary, Alberta
US
₹12.4L PA avg
Avg at Replicon
About
Time management software to track time invested, expenses & attendance
Industries
Software Product
Employee ratings
10 reviews
3.7/ 5
Culture
3.8
Career growth
3.2
Work-life
4.3
Employees rate it well for
Work-Life Balance37Overall Company37
Find them on