Lead AI Developer – Infrastructure Security Automation (L3)

Theomnihire; Mumbai; On-site; INR 3,000,000 - 5,400,000; Full time

Job summary

Theomnihire is seeking a Lead AI Developer – Infrastructure Security Automation (L3) to architect and direct production-grade AI systems, agentic platforms, and LLM-powered security capabilities within its Cyber Security Division.

The role covers end-to-end LLM strategy, RAG design, threat modeling, and engineering leadership across Python/Java backends, with a focus on secure automation of security operations.

Qualifications

• Experience: 6 to 9 years of software engineering & AI developer leadership experience.
• Education: B.Tech or M.Tech in Computer Science, IT, or related technical field.
• Core Tech: LLM APIs and open-source stacks; orchestration & agents; RAG & vector stores; backend in Python/Java.

Responsibilities

• Architect AI systems and agentic platforms for security operations.
• Own end-to-end LLM strategy and cost/latency optimization.
• Design evaluation frameworks for hallucination, latency, and accuracy.
• Lead RAG design with ingestion, embedding, and hybrid retrieval.
• Drive agentic system engineering with planner/executor patterns.
• Establish responsible-AI standards and threat modeling.
• Balance model selection, fine-tuning, distillation, and quantization.
• Define engineering standards and oversee backend services.

Skills

Python

Java

AI Leadership

AI Systems Design

Education

B.Tech or M.Tech in CS/IT

Tools

LangChain

LlamaIndex

LangGraph

Haystack

LangSmith

Langfuse

Arize

Helicone

pgvector

Pinecone

Weaviate

Chroma

FAISS

Azure AI Search

Vertex AI Vector Search

Job description

Lead AI Developer – Infrastructure Security Automation (L3)

Job Title: Lead AI Developer – Infrastructure Security Automation (L3)

Working Hours: 9:00 AM – 6:00 PM

Mode of Interview: Face-to-Face or MS Teams

Position Summary

The Lead AI Developer – Infrastructure Security Automation (L3) is a principal technical lead role responsible for architecting and directing the delivery of production-grade AI systems, agentic platforms, and LLM-powered security capabilities . Operating within the Cyber Security Division, this role leads the design of advanced RAG systems, agentic state machines, model strategy (hosted and open-source fine-tuning/quantization), and evaluation frameworks . The Lead sets technical standards across Python/Java backend services, establishes threat modeling practices for AI workloads, and collaborates closely with security, DevOps, and infrastructure leadership to automate critical security operations

Requirements

Key Responsibilities

AI Systems & Agentic Architecture: Architect and lead delivery of production AI systems and agentic platforms for vulnerability triage, remediation copilots, log/incident analysis, policy-as-code reviews, and natural-language query over security data .

End-to-End LLM Strategy: Own LLM application architecture end-to-end: model selection, prompt strategy, RAG design, agent/tool orchestration, memory management, guardrails, and cost/latency optimization .

Evaluation & Quality Frameworks: Design and implement evaluation frameworks including offline benchmarks, online A/B and shadow testing, human-in-the-loop reviews, regression suites, and continuous quality monitoring for hallucination, accuracy, latency, and cost .

Advanced RAG & Knowledge Systems: Lead RAG design covering ingestion pipelines, chunking and indexing strategies, embedding selection, hybrid retrieval (vector + lexical + re-ranking), and grounding patterns .

Agentic System Engineering: Drive agentic system design using planner/executor patterns, tool use, multi-step workflows, error recovery, and safe action boundaries, while setting up agent observability and tracing .

Responsible AI & Threat Modeling: Establish responsible-AI standards (prompt-injection defenses, PII handling, output validation, red-teaming, audit logging) and lead threat modeling for AI/LLM systems (adversarial prompt attacks, data poisoning, supply-chain risks) .

Model Optimization & Strategy: Evaluate hosted vs. open-source models; direct fine-tuning, distillation, and quantization strategies to balance accuracy, latency, cost, and data sovereignty .

Engineering Leadership & Standards: Own production backend services in Python (and Java where required), define engineering standards (design reviews, testing, observability), and oversee internal tool UIs/dashboards .

Candidate Profile & Qualifications

Experience: 6 to 9 years of software engineering & AI developer leadership experience .

Education: B.Tech or M.Tech in Computer Science, IT, or related technical field .

Core Technical Skills:

LLM & Model Expertise: Deep hands-on experience with LLM provider APIs (Anthropic, OpenAI, Azure OpenAI, Vertex AI) and open-source model stacks (Llama, Mistral, Qwen) including self-hosted serving, fine-tuning, and quantization .

Orchestration & Agents: Strong command of orchestration frameworks (LangChain, LlamaIndex, LangGraph, Haystack); hands-on agentic experience (tool use, planner/executor patterns, state machines) and observability (LangSmith, Langfuse, Arize, Helicone) .

RAG & Vector Stores: Production RAG experience across ingestion pipelines, hybrid retrieval, re-ranking, and tuning with vector stores (pgvector, Pinecone, Weaviate, Chroma, FAISS, Azure AI Search, Vertex AI Vector Search) .

Engineering & Backend: Production backend services in Python (and Java where required) alongside basic frontend/dashboard design .

Mandatory Certifications (Any 1 required):

CISSP, CCSP, Azure Security Engineer Associate (AZ-500), Azure AI Engineer Associate (AI-102), AWS Certified Security – Specialty, or GCP Professional Cloud Security Engineer .

Desirable Certifications:

CISM, CEH, OSCP, GCP PMLE, AWS ML Specialty, CKA, Terraform Associate, or SC-100 .