Senior AI Developer InfraSec Automation (L2)

Job Description – Senior AI Developer InfraSec Automation (L2)

Position : Senior AI Developer InfraSec Automation

Grade/Level DM/M/SM

Reporting to : Lead AI Developer InfraSec Automation

Employment Type: Employee - Full Time

Work Location: Mumbai

Key Focus area : InfraSec Automation

Key Responsibilities :  

• Design, build, and ship LLM-powered features and AI assistants that support security workflows — vulnerability summarization , log triage, remediation recommendations , policy and configuration review, and n atural-language query over security data.
• Develop prompt templates , system prompts, and structured-output schemas; iterate prompt design using offline and online evaluation.
• Implement Retrieval-Augmented Generation (RAG) pipelines end-to-end : chunking strategy, embeddings, vector store integration, retrieval tuning, and grounding.
• Build and operate AI services and APIs in Python that wrap LLM providers (Anthropic, OpenAI, Azure OpenAI, Vertex AI) and open-source models, including tool/function calling and basic agentic workflows.
• Develop supporting backend services in Java where required, and lightweight frontend components ( HTML, CSS, JavaScript ) for AI-driven internal tools and dashboards.
• Implement evaluation harnesses, golden datasets, and regression suites for LLM outputs, track accuracy, hallucination rates, latency, and cost.
• Apply responsible AI and security controls : prompt-injection mitigations , PII redaction, output filtering, rate limiting, audit logging, and access controls.
• Integrate AI services with scanners, ticketing systems, SIEM, and monitoring stacks to deliver actionable, automated workflows.
• Automate AI/ML ops tasks (data prep, embedding refresh, eval runs, health checks) using Python and Shell scripts.
• Assist in testing, debugging, and tuning AI applications for reliability, performance, and cost; maintain documentation, code repositories, and deployment artifacts.
• Collaborate with security, engineering, infrastructure, DevOps, and data teams to translate requirements into AI-driven solutions.

Qualification and Work Experience 

Qualification : BE / BTech (Similar Education Background) with 60%+ throughout the academics. 

Work experience : 4-6 Years 

• 4–6 years of experience in software development, ML/AI engineering, or applied AI, with at least 1–2 years building production LLM-based applications.
• Strong Python for AI/automation work; comfortable with FastAPI/Flask, async patterns, packaging, and testing frameworks.
• Hands-on with LLM provider APIs (Anthropic, OpenAI, Azure OpenAI, Vertex AI) and at least one orchestration framework (LangChain, LlamaIndex, LangGraph, Haystack, or equivalent).
• Practical experience with prompt engineering, structured outputs (JSON schema, function calling), and tool use.
• Hands-on RAG experience: embeddings, chunking, retrieval tuning, and at least one vector store (pgvector, Pinecone, Weaviate, Chroma, FAISS, Azure AI Search, or Vertex AI Vector Search).
• Working experience with LLM evaluation: building eval sets, using frameworks (RAGAS, DeepEval, Promptfoo, LangSmith, or equivalent), and interpreting results.
• Working knowledge of Java for backend services and REST APIs; basic frontend with HTML, CSS, and JavaScript.
• Shell scripting and Linux fundamentals; Git, CI/CD, Docker, and at least one cloud platform (Azure, GCP, or AWS).
• Basic understanding of infrastructure and security concepts (vulnerabilities, patching, logs, IAM, access controls) enough to apply AI in this domain.
• Awareness of AI safety and responsible-AI practices: prompt injection, data leakage, hallucination, bias, and content filtering.
• Good written and verbal communication; collaborative working style.
• Familiarity with SIEM platforms (Microsoft Sentinel, Splunk) and ability to write queries (KQL or SPL) for AI-assisted detection and investigation use cases.
• Basic knowledge of CSPM, cloud security controls (NSGs, WAF, IAM policies, conditional access), and their relevance to securing AI service deployments.
• Understanding of Infrastructure as Code (Terraform or Bicep) for cloud-native AI service deployment and configuration.
• Exposure to data privacy regulations applicable in India — DPDP Act, RBI IT Risk Framework — and their implications for handling personal or sensitive data in AI systems.
• Familiarity with vulnerability management concepts: CVE, CVSS, EPSS scoring, CISA KEV catalogue, and patch prioritisation workflows.

Competencies /Expertise Required (Functional & Behavioral)

Systematic problem-solving — able to decompose AI product requirements, debug LLM behaviour (prompts, retrieval, tools), and apply structured, risk-aware solutions within defined architectures.

• Strong analytical and critical thinking — interprets evaluation metrics, application logs, and security telemetry to drive improvements in AI quality, reliability, and cost.
• Adaptable to a rapidly evolving AI landscape — new models, frameworks, and techniques appear constantly; the candidate keeps pace and applies them pragmatically.
• Learning agility — actively explores new models, prompting techniques, evaluation methods, and AI libraries to lift product quality and security posture.
• Ownership and reliability — sees AI features through from prototype to production with attention to evaluation, observability, and documentation.

Certifications

Preferred / Advantageous:

• CISSP – Associate of ISC² (working towards full CISSP)
• CEH – Certified Ethical Hacker (EC-Council)
• Google Professional Machine Learning Engineer or AWS Certified Machine Learning – Specialty
• Microsoft Certified: Azure Administrator Associate (AZ-104)
• CCSP – Certified Cloud Security Professional (ISC²) — working towards
• Google Associate Cloud Engineer or equivalent cloud platform certification